DISCLAIMER NOTE: I have never received any compensation,monetary or otherwise, from the persons/companies named in these minireviews. My opinions are based on use of purchased software, freeware, or shareware, and are my own. Note: I would appreciate anyone with additional reviews, comments, etc., on commercial/shareware/freeware programs for molecular biology on the apple macintosh to send them to me over e-mail. USING THE APPLE MACINTOSH IN MOLECULAR BIOLOGY (Dec 1990) Note: New version, with new reviews coming out in the first months of 1991. A. General Information 1)Advantages of Macintosh 2)Program Availability B. Commercial DNA and Protein Sequence Analysis Software C. Feature tables, and their use D. The "Big Three" E. Commercial Software Reviews: 1) MacVector (IBI) 2) GeneWorks (IntelliGenetics) 3) MacMoly (Soft Gene Berlin) 4) The DNA Inspector II $295? 5) Gene Construction Kit 6) DNA/RNA Builder and Protein Predictor 7) MacMimic 8) DNA Parrot Gel Digitizer System 9) Plasmid ARTIST 10) Ball and Stick: 11) MacImdat 12) DNASTAR-Mac - DNASTAR-Mac 13) Applied Biosystems 14) STADEN-Mac F. Noncommercial and Public Domain 1)DNA Strider, 2) Mac Sequence analysis programs, Bernard Bellon, 3) PM Sequence analysis 4) Genetic Linkage analysis/Pedigree analysis programs. 5) Mac Plasmap, v. 1.8 6) PLOTA protein analysis package 7) Pearson Sequence Analysis Programs,William R. 8) PROPLASM, plasmid drawing program, P.F. Lurquin, 9) GenBank Search (HyperCard) 10) AUTHORIN (IntelliGenetics) G. DNA and Protein Sequence Database Sources: 1)GenBank 2)Prosite H. HyperCard,FaceWare, and Writing Sequence Analysis Routines on the Mac I. FaceWare E) Journals/Magazines/Publications of Interest J. Online Services, Commercial: 1. BioTechNet: 2. America Online: K. Noncommercial Academic Services(GenBank,EMBL) a) GenBank Sequence Retrieval System: b) GenBank Sequence Data Submission Form c) Online FASTA d)Getting Mac Programs off The Servers: L. Future developments: *Written by Peter Markiewicz, 1989 Dept. Biology, MBI, UCLA, 405 Hilgard Ave, Los Angeles,CA 90024 Bitnet Address: JMILLER%VXBIO.SPAN@STAR.STANFORD.EDU A. General Information This handout summarizes some of the Macintosh programs and hardware of value to molecular biologists, biochemists, geneticists, and others in the life sciences. It is not necessarily comprehensive! Each section cites a few programs and/or hardware that the author has had experience with, and as such constitute "minireviews." If you have questions, opinions, or comments on this document, please send them to me at the above BITNET address. I am especially interested in helping programmers who may just be starting to learn 'Mac', people wanting to port their programs to the Mac environment, and so on. 1)Advantages and Performance of Macintosh vs. mainframes and other Micros. The Apple Macintosh uses a Motorola 68000, 68020, or 68030 central processing unit (cpu) running at 8-50 mHZ, and as such, is quite fast in calculation compared to other microcomputers. There are large speed differences between A 68000 Mac Plus and a 68030 MacIIci. A representative comparison "benchmark" program is the FASTP package of Pearson and Lipman (Science, 10 Mar 1985). This program,written in "C", is designed to rapidly scan the NBRF protein database library for matches to a test protein sequence, and is quite demanding of the hardware of a microcomputer. Results for running this program using a 20-amino acid search peptide on several machines is shown below: Apple II not tried, but would be nearly an hour. IBM PC/XT 10-12 minutes. IBM PC/AT 3 minutes. Macintosh 512K 4 minutes. Mac Plus/SE 3 1/2 minutes Macintosh II (68020) 1 1/2 minutes Macintosh IIcx/i (68030) not tried, should be about 30 seconds for the fastest models. Macintosh II/fx (40MHz) 3-5 times faster than other Mac IIs at most tasks. -386 Box (20MHz) comparable to mac IIx, VAX 11 1 minute. Mainframe not tried, could be a few seconds in some cases. This should dispel claims that the Mac is a slow machine relative to other micros. The Mac offers several advantages over the other alternatives. First and foremost, it is very easy to use. The typical user spends only 20% of the time learning how to use a program s/he would spend on another computer. This fact was made obvious in my previous job.Even with a training course, only one out of several scientists managed to learn the VAX version of Intelligenetics software well enough to use it effectively,while everyone learned to use similar software on their Macs, without training. Another department spent $60,000 in 1987 just to teach their (12) members to use IBM PCs! The availability of Windows 3 in the PC world is making things easier, but you still have to drop into DOS to manipulate files. Windows does not handle printing in the simple(for the user) method used by the mac, and still requires custon drivers for each program. It is somewhat amusing that the same people who used to complain about the mac's "toy" windows/menu interface today are touting Windows 3 as the second coming! On a fast 386, Windows feels a lot like a 1985 512k mac, which was hardly a speed demon. Rather than being a threat, Windows validates the mac interface. Final proof of the victory of the "mac" style of doing things is the recent decision in the Human Genome Initiative to adopt the Mac interface. The ability to produce publication-quality graphics with a Mac plus Laserwriter (bypassing your Art department) is another advantage. With a Mac II , a color drawing program like PixelPaint (SuperMac) and a HP-PaintJet or Tektronix color thermal printer or color laserprinter, one can put out the equivalent of a color magazine from their desktop. 2)Program Availability Until recently, there was little software for scientific applications. In the last year, numerous useful programs have become available. Most of these are best ordered directly from the company, or through a mail-order supply house, because the typical computer store won't have them in stock. Prices range from nothing (for the public domain programs) to $2000-$3000 for some of the bigger packages. Most of the commercial packages are hardware copy-protected,meaning that even if you pirate a copy, it won't work because you don't have a little hardware box attached to your mac. With the new operating system and MultiFinder, it is necessary to upgrade to AT LEAST Mac Plus level. Nobody supports the old 400k Macs while programming anymore! Check any of the Mac computer magazines (such as Macworld or MacUser) for details. Nearly all programs work as well or better on a Mac(+) or Mac SE than they do on a 512K. About 90% of the current Mac programs run without modification on the various Mac IIs. A notable exception are paint-type programs, which may have to be run in a monochrome mode to work properly. Paint programs using full Mac II color are tremendous memory hogs, requiring 2-5 megabytes. Programs which align several DNA or protein sequences at once typically require huge amounts of memory as well (3-5 megs). With the introduction of System 7 around the corner, anyone with a Mac II SE/30 will want to get >=2megs of RAM. This will allow you to use the virtual memory option, where part of the hard disk acts as if it were RAM. If you have an older 68020 Mac II model, you also have to get a PMMU chip. The Plus,SE,Classic,and LC can't use virtual memory, but will benefit from IAC (interapplication communication). This is a system-wide form of the "hot links" previously supported between individual programs. It allows one program to send its data to another without user intervention, writing files and reloading them, etc. For example, one could copy spreadsheet data into an appropriate word processor supporting IAC, and when you made changes in the spreadsheet, the equivalent info in the word processor document would also be updated! Why you Need a CD-ROM Reader If you are going to run sequence analysis involving the sequence databases GenBank and EMBL, you have to buy a CD-ROM reader. The size of the combined databases is approaching 100megs, so you hard disk won't be big enough. Apple and many third-party vendors sell these readers, and a small NEC reader can be gotten for under $500. B. Commercial DNA and Protein Sequence Analysis Software These are programs that allow entry and editing of DNA or protein sequences, restriction enzyme analysis, homology analysis, translation of DNA sequences to proteins, and sequence alignment from "shotgun" cloning trials. There are several commercial programs providing these utilities, as well as some public domain programs. The Commercial Packages: MacVector,MacMolly, and GeneWorks. Below I review MacVector, GeneWorks, and MacMolly. All three of these programs are in the $2000 price range, and are comprehensive sequence analsis programs. That means, they all provide the basics: --technical support from the company for registered users, including free or low-cost upgrades to registered users. --sequence entry from digitizer. --sequence editing,full IUPAC ambiguity code recognition. --restriction maps,open reading frames, look for DNA hairpins,etc. --protein analysis. --database searches of GenBank or EMBL, aligning the test sequence. to hits in the database. --reading feature tables and linking features to the DNA/protein sequence. --detailed graphics that can be saved/printed/copied to clipboard. --ability to "zoom" in on graphics, to look at interesting regions. --program and database distribution on CD-ROM, ability to use CD-ROM databases in searches. Quite a change from a few years ago, when no comprehensive sequence analysis packages existed for the mac! MacMolly and MacVector have been around for about 2 years, while GeneWorks is a new package. This review does not treat two new packages for the mac. One, DNASTAR, is a mac version of the popular and expensive PC program. The other macSTADEN, is similarly a mac version of the STADEN program which have been on mainframes for several years. Future editions of this review will cover these programs as well. Since the programs all provide the "basics", the following distingushes them using other criteria: --ease of use/installation, mac style interface. --ability to create "publication quality" graphics from sequence analysis results. --shotgun alignment of gel sequences to a consensus. --multiple sequence alignment for maximum homology. --ability to read/use feature tables in sequence analysis. --"hot links" between text and graphics and analysis (changes in sequence automatically update graphical representations, and the results of sequence analysis). Note that NONE of these programs support the following, though they will in the near future: --support for site-directed mutagenesis --PCR oligo design, optimization of PCR conditions --Alignment of DNA or protein sequences for evolutionary homology, and analysis of aligned sets of sequences. In the following, a review of each program is given, and a chart comparing the features of the various programs follows. All of these companies are happy to provide you with "demo" software. Demo software works just like the real program, except you can't print or save your results. I strongly suggest you "test drive" all of these programs to find the one you like best. C. Feature tables, and their use: Several of the commercial programs began to support "feature tables" in 1990. What is a feature table, and why are they important? Any DNA or protein sequence submitted to GenBank,EMBL,SwissProt,PIR, or other data- bases contains a section with comments on sequence function, who submitted the sequence, relevant journal articles, important regions of the sequence, etc. The positions of interesting parts of the sequence are summarized in the feature table. For example, If you were to look at the GenBank file for pBR322, the feature table would list the exact positions of the genes for ampicilin and tet resistance. It also lists positions of promoters, operators, terminators, introns, and any other biologically or chemically relevant portions of the sequence. A file for the a protein might list the amino acids in the active site. Obviously, it is often useful to know the locations of these regions if you are going to work with a particular DNA or protein. For example, imagine you want to get the amino acid sequence of a eukaryotic gene, and you only have the DNA sequence of the whole gene region. This requires (1) knowing where all the exons and introns are, and (2) carefully selecting them from the file. Obviously, if you select one base too many, the translation will be incorrect. In the past, these manipulations are quite difficult to select the right region of DNA in your editor--but no more. Several of the commercial programs not "read" the feature tables. Selecting a gene,intron,etc., in a larger sequence is now simply a matter of clicking on its name in a list. I predict that reading feature tables will become universal, and how well a program uses them will be a major consideration in buying it. D. "The Big Three" The following compares GeneWorks, MacMolly, and MacVector. DNA Inspector is also discussed, even though it is not in the class (price and performancewise) of the other programs. 1) International Biotechnologies, Inc MacVector Sequence Analysis Package, Subs. of Eastman Kodak Co., P.O. Box 9558, New Haven, CT 06535 $2500.00 This is not the old Pustell port. This is a completely new program, written by a mac programmer, The new upgrade for 1991 continues the tradition of quality, power, and ease of use. IBI obviously listens to user's complaints, and did a lot of work to give them what they want. They should be congratulated for this! If you have the old MacVector, a very reasonably priced upgrade is being offered in 1991. The program does not run on a Mac Plus, requiring a Mac SE,Classic, or any Mac II. The reason for this is the hardware copy protection scheme, which requires that you have one of the new ADB keyboards. A little box plugs between your keyboard and the Mac. This is a good scheme for protecting the program, since it doesn't require passwords, or inserting a master disk each time you use the program. Unlike GeneWorks, MacVector runs well, is somewhat slowly, on a Mac SE or Classic. But if you are buying a computer at the same time as the software, you will be better off with a MacII LC , si, or f/x than a Mac Classic. Online Help is found under the "About..." menu option under the Apple Menu. It has a simple organization and is reasonably informative (though Help in IG's GeneWorks is more comprehensive). The part most people will be looking at is the setup for a gel reader (if they have one), and setting up the GenBank or NBRF database so MacVector will find them. The program's written manuals are complete and not difficult to use. When you start the program, it presents you with the option of loading a DNA or protein sequence. A large variety of sequence formats are supported, including BIONET/IntelliGenetics,GenBank,EMBL,Staden,Line, University of Wisconsin package format, and plain text. The program recognizes the standard IUPAC codes for ambigous DNA/proteins. The IBI program also a supports sonic digitizer for direct input of sequences from a gel into an editor window. You need a modem between the digitizer and the Mac to make it work. It is possible to use non-IBI digitizers with the program,if your digitizer allows you to adjust base positions.MacVector provides powerful text editors for your sequences. I did not have the chance to check this feature, lacking a digitizer, but reports from others have indicated that it works well. You can also configure your numeric keypad on the keyboard for sequence data input. Once the sequence is loaded, it appears in a text editing window. The window lists sequence length, and any regions you have selected are also listed. The response of the editor is a little slower that a word processor, but is quite acceptable. One very useful thing you can set is the number of bases/amino acids displayed before a space is inserted. A sliding scale on the sequence entry window allows you to vary between 1 and 10 bases per block. The interface of the sequence editor, and everything else, is pure mac. IBI has rigidly adhered to Apple's user interface guidelines for writing a mac program. Thus, file loading is under the file menu, and editing restriction enzyme/protease specificities, mobility standards,and homology scoring tables is under the Edit menu. I find this approach easier to work with than the poppup object-oriented approach taken by GeneWorks(see below). Cutting and pasting to the ClipBoard is supported for text and graphics, and MacVector is MultiFinder-friendly. This program receives a 100% rating on Mac interface! I was very impressed with how easy it was to operate this program. Anyone who has used a mac for word processing/drawing should be able to use this program right away, without looking at the manual! This is not true for MacMolly or GeneWorks, and only partly true for the DNA Inspector. The programs' speed is impressive as well. When you you point and click, things happen! In the other programs, there is less immediate user feedback. To get additional information about the sequence, you click on little icons on the sequence editor window(s). One locks/unlocks the sequence (very similar to DNA Strider,but the program doesn't explain why you can't type into a locked file,which Strider does). Other icons create windows connected to the editor window. The most powerful feature of the MacVector editing windows, in my opinion, is the links to the comment and feature table windows. Clicking on a listed gene region selects the feature in the sequence itself. The selected region may be copied/pasted to a new editor window. If you are working with a gene sequence with several introns, this GREATLY simplifies getting out the coding sequence! Any analysis can be confined to the selected feature. For example, you could confine a codon frequency analysis to one intron of a longer sequence without having to put it into a new file. To do an analysis of the sequence, you select one of several routines from the appropriate menu. A dialog window comes up, usually with dozens of parameters you can set for the analysis. The complexity of the analysis options is somewhat confusing at times. For example, in a restriction site analysis, it is possible to confine the cutting to certain regions of the sequence. They must be selected in the dialog in order to make the OK button function. It might be good to set default values in the analysis dialogs so that the OK buttons are always clickable. The results of an analysis appear in several windows with text and excellent graphics, any of which may be printed or saved as a file. The graphics windows may all be "zoomed", meaning you can select a small region of the output, and have MacVector enlarge this region. This ability has become standard in Mac analysis packages, as it is found in MacMolly, GeneWorks, and to some extent in DNA Inspector. (I might note that the original "Zoom" windows date from a 1985 program called MacGene). The results are also presented as text with the actual numbers used to create the plots.These are useful if you want to load the numbers into a spreadsheet for further analysis, or another charting program like Cricket Draw. An extremely powerful feature of MacVector, again related to features, is the "annotated sequence" output. Most of the analyses allow you to use this option. The best example of this is in the restriction mapping function. If you select "annotated sequence" as one of the numerous options in the dialog, you get a window with the sequence printed along with the positions of the restriction sites. So far, this is quite similar to DNA Strider. However, MacVector also marks all the listed genes,promoters, introns, etc., as well. This means you can look at your output, and immediately tell whether a particular restriction digest is going to cut a coding sequence in half, preserve a termination signal, etc. This type of information is essential to vector construction. You can also generate a "graphical" annotated map". This means that the program draws a picture of your sequences (better than I can draw), with both restriction sites and features listed. It actually draws two pictures in the window, the top one with the restriction sites, and the bottom one with all the features listed. The size of blocks, color, etc. can be configured under the Options menu. You can zoom on it with the magnifying glass. The picture is as good as any you would find published in a professional journal! This window is not quite as sophisticated as the homologous one in IntelliGenetic's GeneWorks (see below). In GeneWorks, you can actually draw on this picture, using a set of MacDraw-like tools. In MacVector, you will have to import the picture into MacDraw to prepare it for publication. When using these output windows, I felt I should be able to select the objects within them, and have the corresponding sequence selected. If would be very useful,for example, to be able to select two restriction sites, double-click between them, and have the sequence selected in the MacVector sequence editor. Of course, you can enter a particular start and end of a restriction site into the feature window to do this. MacVector has a large battery of routines for analyzing protein sequences for structural features, called the "protein toolbox". It contains routines for analyzing hydropathy, secondary structure,surface probability, flexibility, and antigenic index. Once again, the program creates the results as typical plots, files with text results, and a graphical depiction of the sequence with features listed. The graphs may be zoomed on with the magnifying glass. These are all good algorithms, but it might be useful to include a few more, such as routines searching for leader sequences and membrane-spanning peptides. MacVector has two routines for analyzing sequence homology. The first is a dot-matrix program, allowing you to compare 2 DNA or protein sequences. This routine creates the typical dot-matrix plot, and an alignment of the sequence in text (in another window). You can use the magnifying glass icon to zoom in on interesting regions of the dot-matrix. Amazingly, the aligned sequence window then shows you the local alignment! Though the other packages have this feature, MacVector's is the best in my opinion. There is also a FASTA-type algorithm for searching sequence databases. The routine can also search the sequence comments and features for comments for author or title names. Wild card searches (*.*) are supported. The search routines are now a joy to use. You get a very maclike window to type in the search word for comments. In a matter of seconds, the list of files containing your search are available, and clicking "Browse" allows you to look at the comments. You can also extract the file with the sequence itself at this point for further analysis. Note that an actual search for a particular sequence author, etc. can take a long time. The homology search can take several minutes looking for a protein in NBRF, and an hour going through GenBank. This is true for all programs. Note that the search is faster if you have a hard disk big enough to hold the contents of the CD-ROM database provided in the program. After a search is complete, all file names are listed, and you select them by clicking. The program lists scores, and displays the alignment, all in separate windows. MacVector does not currently support two more complex forms of alignment, which many molecular biologists require for their work. The first is alignment of gel sequences into a consensus. Currently, only MacMolly and the new Mac version of DNASTAR support this. You could use the MacVector dot matrix routine to align a small set of gel sequences, but this is not a true shotgun alignment that builds a consensus, but a series of pairwise alignments. Hopefully, "shotgun" alignment will be included in a future version of the program. Another thing that might be helpful would be to check GenBank sequences for file corruption. Using a checksum, this should be trivial. It is justified because many users get their sequences electronically from GenBank, over often noisy phone lines. Another form of alignment users are interested in is aligning several sequences, known to be evolutionarily related, all at once. MacVector and GeneWorks do not have true multiple sequence alignment ability. This is an important feature that should be implemented in a >$2000 program. Currently, one must use programs like CLUSTAL on the VAX, or a small number of public-domain programs for this purpose. I have received literally hundreds of requests for this type of multiple sequence alignment. Comeon, there's a market! To conclude, check out the chart at the end of this section to see how MacVector rates compared to the other programs. It is no longer the ONLY package for all users, but is for many. It is best for those new to sequence analysis. It is a good choice for anyone with a "low-end" macs like SEs and Classics. It is still the only choice if you don't like reading manuals. If you are running a big sequencing project that requires shotgun alignment and consensus building, MacMolly is the better choice. And if you are looking for an incredibly full-featured, Cadillac of a "Sequence analysis environment" for a Mac IIsi, ci, or f/x, you should check out GeneWorks. 2) PC Gene/GeneWorks--Macintosh Version: IntelliGenetics 700 E. El Camino RealMountain View, CA 94040 (415)-962-7300 When GeneWorks is launched on your computer, you know that something unusual is going on. For one thing, it takes it over a minute to load. The enormous size of the program itself (1.2 megs) also indicates that it is different from earlier programs like MacMolly and MacVector. GeneWorks represents an attempt to move beyond traditional sequence analysis programs. In the typical program, you work in a "proceedure" mode. You enter the sequence in a text editor. Then you select an analysis option. The results of the analysis are put into a window or file. There is little connection between the sequence, the algorithm, and the output data. Not so in GeneWorks. This program is part of the wave of "object oriented" sytems that will appear on computers everywhere in the next few years. In GeneWorks, your sequence, the analysis routines applied to it, and the output from them are all linked into a single "object". This is formalized as being various "views" of a sequence. The idea is that all sequence analysis is a way of viewing a sequence. Looking at the bases in the editor is only one view, and other views have equal ranking! For example, say you load a sequence from a file. Rather than running routines on it, you specify which routines will be applied to the sequence. Every time you edit the sequence, the program quickly preforms all the analyses on the altered sequence! All the views you have defined, which would be output windows in other programs, are updated! One can see how programs like this will be natural when parallel computers become common in the next few years. An example of this coupling is in the feature tables (as discussed for MacVector). In GeneWorks, a window lists the features from the file. In GeneWorks, however, the features are visible in all views. If a protein has a sulfhydryl bridge between two cysteines, for example, a little line connecting them is visible in the graphical view of the sequence. However, it is also visible in the sequence editing window as well! This is a true breakthrough. This is something I have wanted in a computer program for a long time. Imagine editing a sequence, and having all the restriction sites displayed automatically above it, updated if the sequence change knocks out one of them! Thisis truly the future of sequence analysis. GeneWorks takes the concept of linking features to sequence and output to its logical conclusion, well beyond that in MacVector. The following describes a typical session with GeneWorks. The program comes with two manuals, a quick tutorial, and a complete (huge) manual. You will need to go through the tutorial to get a feel for the program. The big manual covers sequence analysis and additional topics, like preparing graphics for use in posters/publications. When GeneWorks is started and a sequence loaded, a window appears showing your sequence with an icon, in a manner reminescent of the Finder (This is a bit redundant under MultiFinder). Clicking on it gives an editor window similar in many ways to MacVector.(GeneWorks also supports gel digitizer input of a sequence). Features are directly listed in the editor window (for example, if two cysteines are linked by a sulfhydryl, a little |-S--S-| actually links the two amino acids. You can see exactly where features like restriction sites or sulfhydryl bonds are during editing. In MacVector and DNA Strider, these features appear in a separate, noneditable, "annotated sequence" window. Once the sequence is loaded, you can look at it using the variety of "views" listed under a pulldown menu, including graphical, text, feature tables, etc. GeneWorks goes beyond MacVector in the sheer number of ways a sequence can be looked at,the COMPLETE consistency of defining a sequence analysis as simply another "view". The font,style,size,color, etc., of EVERY view window may be modified independently of the others, and many of them have a set of MacDraw-type tools for adding to the picture! The appearance of a sequence feature may be modified. You can also modify them in MacVector, but GeneWorks provides more options. Double- clicking on a sequence feature brings up an editing window in which the graphic appearance of this feature can be controlled. Other view windows allow you to look at the sequence, as recorded in the computer file (what a text editor would show you). You can see the feature table information in a separate window, as in MacVector. The analyses window has the results printed in text form, in a spread- sheet. The sheer number of ways of interacting with your sequence is astounding. A casual user will never appreciate the full power in GeneWorks, and will probably find it more confusing than MacVector. But anyone who does a lot of sequence analysis will end up using the program for more and more analysis, involving things that simply can't be done in other programs. An example of using these options follows. If you have a sequence in a text editor, you can have windows connected to it showing tables of open reading frames, restriction sites,and other analysis routines. You also have the graphics window, which can be toggled to show any of the table results in graphic form. You can draw lines, boxes, etc., around interesting parts of the graphs. If you double-click on any table entry in any of the windows, the corresponding region of the sequence is selected in the sequence editor. As the sequence is edited, changes can be reflected in the other windows. To set up an analysis, you click on the control panel icon on the sequence editing window. You specify which routines will be applied to the sequence. The first time you do this, GeneWorks takes several minutes to calculate the analysis for the sequence. After this updates are automatic. A large set of algorithms are supplied. I particularly liked the open-reading frame (ORF) analysis, which allows you to look only for ORFS with a Shine-Delgano sequence. One unique feature of GeneWorks is that you can actually modify the algorithms used in sequence analysis. If you select an algorithm and press the New button, you can change constants, cuttoffs, etc., in the algorithm. Other programs let you modify some of these values, but GeneWorks is much more consistent and complete in this respect. The feature will benefit "power users" more than casual users. Such deep access into the innards of the program is possible only because it was written in an object-oriented fashion. During analysis, the program creates a window with the results. As mentioned above, these windows have drawing tools in them, and are linked to the sequence text. You also have a magnifying glass for zooming to interesting regions of the analysis. Each time you add an analysis, its chart is dropped into the sequence window along with those already there. It is possible to drag the charts around, and superimpose them on top of each other. If that weren't enough, there are other subviews within the graphics window. You can look at the features, ORFS, restriction maps, plots, or notes on the graphics. You may select objects in a graphics window, and copy/paste the sequences they cover. To do this, you select while holding down the option key. If you do this to a restriction map, the fragment between two sites is selected. You can then copy this to a sequence editor window. You can make a picture of the gel pattern expected for the particular restriction enzyme(s) choosen. You can select one of the bands, but I wasn't able to copy the sequence represented by the gel band to an editor. This is a break in consistency which should be fixed. GeneWorks has a big set of analysis routines for protein sequences, which are invoked in a fashion "concurrent" similar to DNA analysis. If you have a large number of calculations selected, it can slow down editing. You probably will want to have them run in the backround while editing. GeneWorks has a completely unique feature, which is the ability to examine the result of two analyses for correlation. For example, one can plot the results for hydropathy against the results for secondary structure. You get a window with a scattergram showing where high hydropathy correlated with high sheet-forming propensity. This is something I have wanted to do for a long time, and GeneWorks is unique in providing it. If you want to do this type of "analysis of analyses", you will have to get GeneWorks. Like the other programs, GeneWorks can do database searches. You will need a >100meg hard disk or CD-ROM reader to use it effectively. The database searches are the most confusing part of GeneWorks. You set up a search using the same tightly interlinked nest of windows as elsewhere. To run a search, you first must load a database. You do this not by simply opening a database file, but by creating a "Query" and loading the file. You then use the "browser" functions to select a sub-database as a search. The sub-database is a query. This allows you to have a series of virtual sub-databases to work with without physically copying out the sequences into a new file. The query lists the sequence names, and you can get the actual sequences in the databases simply by clicking on them. So far so good. However, I was stumped for over an hour, even with the documentation sent with a demo disk, on how to search the database or sub-database with an individual sequence. You must select the search algorithms used with the databases in a fashion analogous to select analysis routines for a sequence. You then run the algorithm. After it runs, the results are invisible! You have to first (1)select the sort by.. poppup menu, and (2)scroll the horizontal scroll bar in the query sequences (database) window to see the column with the scores beside each sequence. Only after this is done can you look at your hits! This is quite confusing to a novice, who might never learn how to do searches. If you manage to figure out the database functions, however, they are extremely powerful. You can use the "browser" functions to restrict the searches so they only include sequences from particular species, enzyme types, keywords, or authors(!) This can greatly reduce the time necessary for a search. You then click on the now familar control panel icon to select search algorithms, exactly like selecting sequence analysis algorithms. Again, GeneWorks is uniquely powerful in this respect. For example, you could search for a particular pattern of hydropathy or secondary structure rather than sequence. No other program allows you to do anything like this. Of special interest is the provision for "hooks" within the program. This means that third-party software could be linked into the main program, and function as part of GeneWorks. This is an extremely powerful feature, similar to HyperCard XCMDS and XFCNS. This means that if you buy GeneWorks, you may be able to buy only the analysis subprograms you really need as they become available. It is possible that a small third-party industry will spring up around GeneWorks, if IntelliGenetics lets other people write routines to link into GeneWorks. This again, is a new, unique feature to sequence analysis (don't tell me about DNASTAR--the programs are not part of an "hot linked"environment like GeneWorks). From the above, the reader must be convinced of the novelty and sheer power of this program. Nothing like it has ever appeared on any computer, though the $50,000 StrataGene package from a few years ago had similarities. The list of features unique to GeneWorks is impressive. If you are a mac power user, this is probably the program you want. I think that GeneWorks is groundbreaking. In time, sequence analysis on any computer will be in a similar "environment". If you like to be on the cutting edge, GeneWorks is for you. However, if you have a lab with several novice computer people and one mac SE, and you want to do the basics on your micro, you are better off with MacVector or MacMolly. Now for the problems. Like MacVector, GeneWorks has no provisions for true multiple sequence alignment, or shotgun sequence alignment. One hopes that future versions of the program will have these essential components of sequence analysis software. A serious consideration is the complexity of the program. GeneWorks is lets you do an enormous number of things. It is a sequence analysis and desktop publishing program in one. It allows you to compare the result from conventional analyses to each other. A new user can become lost in such complexity. Even though GeneWorks is an order of magnitude easier to use that a PC or mainframe program of similar power, it is not very intuitive. I was able to run most of the features of MacVector in a few minutes, without consulting the documentation. By constrast, there were portions of GeneWorks which were not easy to understand, even for an experienced user like myself. With the exception of the homology search, it was worth the effort to learn the program's details, which are consistent, if novel. You will be doing a lot less menu selecting and a lot more pushing buttons in windows when using GeneWorks. The power in GeneWorks has come at a performance price, at least for today's macs. Many operations in the program take a long time to happen. Simply loading the program, for example takes over a minute. The program is not slow during analysis, but the fact that analysis may run concurrently with everything else slows things up. To run GeneWorks, you will need at LEAST a MacIIci/MacIIsi with 2 or more megs of RAM, and a macII f/x (40mHz!) would be ideal. A Mac SE,LC, or Classic simply don't have the guts to run GeneWorks! Finally, I note that this is the first release of GeneWorks, and programs of this complexity always have bugs on first release. The use of object- oriented programming has probably minimized this, but there are problems. The program sometimes runs out of memory (on a Mac IIsi with two megs of RAM). With more than 4 megs, the problem disappears. There were also annoying minor problems, like the cursor not changing from a text selection caret to an arrow when necessary. Sometimes trying to select a pulldown resulted in the watch cursor spinning for several seconds while GeneWorks did some obsure calculation. Another example: On invoking a gel picture of a restriction digest, I got the same warning message twice. I then double-clicked on the sequence icon, and the program instead brought up the gel picture window again, with the same(!) warning. That ain't a feature, that's a bug! Some commands are primitive. The Find command, for instance, is too simple to be useful. Like MacVector, GeneWorks does no checks for file corruption when reading a sequence. To summarize, this program is a very serious competitor to MacVector in its first release. It is more powerful and comprehensive, but requires an equally powerful mac for running, and possibly a power user to take full advantage of its unique capabilities. 3) MacMoly, Soft Gene Berlin, Offenbacher Str 5, D-1000 Berlin 33, Te.# 030/8211407. Write to this address for pricing and order forms. MacMolly is a comprehensive program offering many useful routines, a sequence database, and a somewhat quirky user interface. The 1991 version of the program has a considerably improved sequence database. Analysis is done by 4 programs, dividing the functions of sequence editing and analysis. The individual functions are quite good, and some like gel sequence alignment/contig are currently unique to MacMolly. The program is a little behind the times, however, in that it can't read feature tables, has limited graphics, and does not use sequence files in other than MacMolly format. Converting a file to MacMolly format is similar to that in DNA Inspector. The four programs in MacMolly give the user the ability to enter the sequence, analyze for the standard features, translate it to a peptide and analyze it, and do shotgun sequence alignment of multiple sequences. The shotgun sequence alignment is the best of any package on the mac. If the main thing you want to do is to align a lot of sequences to a consensus, this is the only program for you. The program generates the consensus, and shows the alignment of the individual sequences. Unlike DNA Inspector, MacMolly is not limited to 10 sequences, or 10k file lengths. In fact, it is a good replacement for STADEN or GCG in this respect. Until the other commercial programs support this feature, MacMolly remains the program of choice for those embarking on big sequencing projects. This is despite the somewhat strange arrangement of the interface. In MacMolly, the user frequently has to select one menu in order to get another menu. Apple's user interface guidlines clearly state that menus should never call other menus, and this makes MacMolly a more difficult program to learn than MacVector or even GeneWorks. I have found that people who use the program either love or hate it. The ones who hate it are usually longtime mac users who dislike the unconventional way of doing things in MacMolly, the weak graphics and printing, etc. The ones who like it invariably have a shotgun alignment project going. Sequence Analysis in MacMolly creates the usual set of multiple windows with output, which are zoomable. MacMolly's output graphics were the first to incorporate the "zooming" feature in MacVector and GeneWorks. Much of the output is text rather than pictures. The results can be saved or printed. I have found that the pictures produced by MacMolly are crude compared to GeneWorks or MacVector, especially when printed. The inability to read feature tables is a big problem if you are working with preexisting sequences, but is not so important if you work with all new sequences. Text output can be saved or printed, and the graphics can be zoomed. If you don't need fancy pictures of vectors or restriction maps, this is not a problem. If you primary intent is to do vector construction, you are better off with Gene Construction Kit (see below). Like the IBI and IntelliGentics programs, MacMolly offers EMBL/NBRF database searches by sequence,ID code, citation, or Keyword. The user interface is less Maclike than the new IBI program, but better than the one in GeneWorks. To use the literature searches you will need either a large hard disk, or a CD-ROM reader. The new version of MacMolly no longer uses the clumsy database system, but contains new programs for this. The new programs greatly improve my opinion of MacMolly. To conclude, I find the MacMolly analysis package to be a good overall buy, especially if you have a large shotgun cloning project. The program set has the best multisequence alignment routine a available on the Mac. It is the only program with a good shotgun gel alignment and consensus building routine. It also has good routines for protein sequence analysis. It is somewhat deficient in user interface and graphical output, but provides a large set of useful routines. The zoomable graphics are excellent. If you need shotgun alignment, you will have to get this program (or the new version of DNASTAR)--MacVector and GeneWorks don't provide it yet. 4)The DNA Inspector II $295? TEXTCO, 27 Gilson RD, West Lebanon,NH 03784 (603)-643-1471. The DNA Inspector was written by a molecular biologist (Bob Gross) in compiled Microsoft Basic. The current version of the program works fine on the Mac SE and the Mac II. A new version of the program, with major rewrites of most functions, an improved shotgun gel alignment routine, and several protein sequence analysis routines, became available in summer 1989. This program costs considerably less than MacVector or GeneWorks, but it is a much less powerful program. It will do most basic sequence analysis functions, and may be a good choice for those with only a little money to spend and a low- end mac (plus,SE, or Classic). The package consists of several programs and data files required for the programs to operate. Note that you can't just copy the programs to your hard disk! Follow the manual's directions exactly! The disks come in a plastic dust cover, and a large and well-illustrated manual is also provided. This manual is well-written, and contains a "sample session" tutorial, as well as an appendix describing the algorithms used by the programs to make its calculations. The documentation is one of the best parts of the program. Overall, the package is quite easy to use, even for a novice. The program controls your choices at each stage so it is fairly easy to run. However, the price for this is "modal" operation. Unlike MacVector,GeneWorks, or MacMolly, the DNA Inspector has fixed-sized windows that can't be moved. When the program is running, you even can't click outside the windows to do anything (which makes it useless under MultiFinder). To install this copy-protected program, you must use an "Installer" disk that copies the program to your disk. If you just copy it to your disk, it won't work. You can install the program three times to three disks. Backups are available from the company for $25 The program provides an Import function which allows plain TEXT files (of up to 32,000 characters) made with a text editor or downloaded from another computer to be converted into DNA Inspector files. The internal format of the Inspector's sequence file is unlike any other program on computers today (random access?), and cannot be used by other sequence analysis programs without converting. Needless to say, DNA Inspector can't use feature table information from GenBank or EMBL files. It can convert GenBank files, but not EMBL or other formats. You can strip the comments from these files yourself in a text editor, and then load them. The DNA Inspector sends output to text files, DNA Inspector files, or printers. If you have a color printer, such as an ImageWriter II with a color ribbon installed,the program can print your output in color. Some of the more complex search algorithms, such as dot-matrix homology, are written in 68000 machine language (CLR libraries?), but most are in MS Basic. This makes the program both very fast and slow. It is very fast when doing some comparisons, but chaining between different parts of the program is slow even on a hard disk. In use, I have found that the slow operation is not much of a problem during the analysis, but is when one is moving from one analysis routine to another. The DNA Inspector uses some of the Mac user interface, but nowhere near the level of MacVector or GeneWorks. Its modal operation is similar to PC sequence analysis programs with menu trees. For example, most mac programs allow the user to save/print results AFTER an analysis has been done, and they are loaded in a window. In the DNA Inspector you must specify in advance whether to save/print your output. This is definitely not the Macintosh Way! Its analysis routines offer most of the routines found in other programs for DNA. Protein analysis is not supported, except for simple hydropathy plots. Graphics are restricted to the size of the windows on the screen, unlike other programs. They are useful, but are a far cry from those found in the more expensive programs. The graphics can't be copied to the ClipBoard, or saved as PICT files. The program does one more sophisticated graphic in dot-matrix homology, where one can zoom on an interesting region. It doesn't create anything like the annotated sequences in MacVector or GeneWorks, (or even DNA Strider) where restriction sites and other features are listed on the sequence itself. The Inspector does not support search sequence databases with your sequence. If you need this function, you will have to get one of the more expensive packages. One feature that the Inspector offers that isn't found in MacVector or GeneWorks (but is in MacMolly) is shotgun alignment of multiple DNA sequences from cloning projects. The old version of this alignment didn't work very well, but the current program is much better. Only 10 sequences may be aligned to a consensus, however. The consensus which is generated is good, and lists mismatches at specific nucleotides where they occur. If you only need shotgun alignment, and can't afford MacMolly, this is the program for you. To conclude, the program has some serious shortcomings, but is useful if you are not looking at a large number of sequences, and your emphasis is on DNA rather than proteins. If you are searching databases or analyzing proteins, or want a true mac-style interface, you are much better off with the higher priced packages. If you are a grad student who is using an "old" mac in the lab and can't afford expensive new hardware or software, this may be the best program for you. SUMMARY: WHAT THE BIG COMMERCIAL PROGRAMS CAN AND CAN'T DO These are my subjective ratings, Super>Great>Good>OK>Poor ; No means program doesn't do it at all. MacVector MacMolly GeneWorks DNA Insp. 0)Price ~$2500 ~$1500 ~$2500 ~$300 1)Sequence from Good OK Good No Digitizer 2)Sequence Editor Great Good Great OK 3)Voice Playback Good Good Good Good 4)Features in Great No Super No Editor 5)Read Feature Great No Great No Tables 6)Link Features Super No Super No & Editor 7)Link Features No No Super No & Output 7a)Make Great Good Good No Annotated Sequence 7b)Select No No Good No sequence from graphics 8)DeskTop Great Poor Great Poor Publishing Graphics 9)Zoom-able Great Good Great OK(dot matrix) Graphics 9a)Print Great OK Great OK Graphics 10)General Great Great Great OK DNA/Protein Analysis 10a) Compare No No Great No analyses to each other 10b)Modify Good Good Great OK analysis constants, etc. 11)Search DNA Super OK Poor No & Protein database 12)Search data- No No Super No base using hydropathy or other analysis 12b)Align to Good Good Poor No database 13)Align No No No No multiple sequences 14)Shotgun No Great No OK Sequence Alignment 15)Help Window Good No Good Good 15a)Genetic Great Good Great No code editing 15b)search Great No Great No matrix editing (FASTA search) 16)Help for No No No No Mutagenesis 17)PCR Oligo No No No No Design 18)Speed of FAST FAST SLOW VERY SLOW Operation 19)Robust Great Good OK(1st release) OK (few bugs/ crashes) 18)Minimum Mac MacSE,MacII Mac Plus Mac II,>2 Mac Plus Configuration preferred CD-ROM or megs RAM Any hard (for reasonable CD-ROM or >80 meg hard CD-ROM or disk operation) >80 meg hard disk >80 meg hard disk disk 19)Manuals Great Great Great Great (essential) (essential) 20)"User" requirements Novice Some mac Some mac Novice exp. exp. The typical user for each of the above: MacVector: Lab has a MacSE or MacII of any type, have CD-ROM, planning to do general sequence analysis, no consensus building from gel sequence, users aren't "power users", want program that is easy to run, don't want to read manuals. Lab also has drawing program like MacDraw to "finish" graphics produced by MacVector. MacMolly: Lab has a MacSE or MacII of any type, have CD-ROM, primary use for computer is to manage data from a big sequencing project, users make their graphics in other programs. GeneWorks: Lab has a MacII with lots of memory (>=2megs), f/x is ideal, at least one user plans to do extensive analysis on sequences, compare and analyze the results from sequence analysis. Lab wants to maintain a complete "sequence analysis environment" where everything--sequence analysis, results, prep. of graphics for publication, etc., are in one integrated system. Lab isn't doing a big sequencing project or aligning multiple sequences for evolutionary homology. DNA Inspector: Lab doing only basic analysis, like restriction site mapping, using old mac (Mac Plus or below), and doesn't have money for big commercial programs. Lab may be doing a SMALL sequencing project requiring M13 shotgun alignment. ************************************************************************** Gene Construction Kit-- a new type of Sequence Analysis The following describes a program that is so innovative that is has to be put in a section distinct from more conventional sequence analysis programs: 5) Gene Construction Kit Textco, 27 Gilson Road, W. Lebanon, NH 03784, (603)-643-1471 The program provides CAD (computer-aided design) for designing DNA molecules. It is like a super plasmid draw program, where when you manipulate drawings of your DNA and its genes, promoters, etc., you are manipulating the real sequences as well. So, for example, you could call up a sequence you want to insert into pBR322, label its interesting spots, add linkers or blunt-end it, and clone it into pBR, all onscreen! The "construct window" can show restriction sites, open reading frames, or other sequence features. Clicking on them allows you to select that portion of the sequence. You can scale the graphics to large or small, in order to see a large sequence onscreen. It is possible to control the display so that only subsets of the defined genes and sites are actually visible. There are two views of a DNA sequence, either the graphical one discussed, or as editable sequence. One remarkable feature of GCK is that all the restriction sites, ORFs, and so on remain defined in the sequence editor. For example, if you colored a gene blue in the graphics, the text editor wil show that region of the sequence in blue. Cuts or pasting of sequence done in the editor will be reflected in the graphic picture. To do a cloning operation, one simply cuts or copies a DNA sequence from one window, and pastes it into another window. The receiving DNA is linearized using a restriction site cut. The Construction Kit puts up a window showing the points of ligation at the nucleotide level. You can adjust the ends so they are blunt, add linkers, etc. When the cloning operation is complete, the new composite sequence will show all its new features, while retaining its "history". To access this history in complex vector construction projects, the program maintains a window which shows the steps you used to make the DNA construct, and a "layout" window which provides desktop-publishing capabilities for the plasmid pictures. The program has a wealth of features which make CAD of DNA constructs practical. It offer features that are new and valuable, and have never been available in ANY computer program on mini,mainframe, or microcomputer. I feel it represents the first of new programs which start treating and analyzing DNA and protein sequences in a manner similar to the way molecular biologists actuall think about them. Much of the difficulty of using current sequence analysis software is that the sequences are treated as text, similar to word processor documents. All the information that is derived from the sequence cannot be associated with the sequence, or used in further analysis. The Gene Construction Kit is a bold step towards incorporating our knowledge of what a DNA or protein sequence is for with the sequence in a useful way. I strongly advise any laboratory which is doing vector construction to get this program. Don't wait for the PC version! Apparently, after the program was demoed, an IBM representative contacted Textco about the feasability of porting it into a PC environment. After study, it appears that it would require AT LEAST A '486 PC running OS/2. The highly graphical, object-oriented interface is too much for MS Windows, though the new release is supposed to be much faster. Certain things would make a good package unbeatable. The most useful would be a query system which allowed you to search a vector database for the plasmid meeting your cloning criteria. For example, one could specify a search for all plasmids with T7 promoters, ampicilin resistance, which were less than 6Kb in size. The expanded GCK would list those in its database, and they could be directly selected from the list. A second useful feature would be to put more sequence analysis routines into the program. Right now, all you can do is search for restriction sites. It would be very useful to find open reading frames in a sequence, and then be able to define a subset of those as "genes" in your construct. More complex searches, such as a search for a sequence GANYYCPC with ambiguity would allow online definition of potential promoters, splice sites, and so on in the construct. This would be valuable since cloning operations often create new promoters, terminators, nuclease hypersensitivity sites, and so on. Continued addition of analysis routines would make GCK into a sequence analysis "environment" similar to the one that molecular biologists actually use in their work. Finally, it would be spectacular if GCK could read feature tables from GenBank and other databases. Currently, the program can import files in these formats, but the information provided in the comments about the locations of genes is lost. I have discussed this with Textco and they say they are waiting for the redefinition of feature tables in GenBank and EMBL before they write the routines to recognize them. The price is about $500 for academics. If this seems steep, consider that IntelliGenetics provided a somewhat similar system on a $30,000 Xerox Lisp workstation, with $50,000 for the software! This 'expert' system, called StrataGene, was slow, cumbersome, and was not worth the learning curve for the molecular biologists at my installation. The Gene Construction Kit, by constrast, takes only a little effort for the average Mac user to learn. Its ultimate utility will depend on whether it offers enough advantages for molecular biologists to use it instead of the traditional pencil and paper cartoons. CAD Gene Genetic Technology Corporation P.O. Box 1179 Kendall Square Cambridge, MA 02142 800-462-7179 508-879-0258 Price ???? CAD Gene is another program for computer-aided design of DNA vector and other constructs. Like Gene Construction kit, it allows managing nucleotide sequence and associated site data in a interactive graphic representation. It links graphic representations of restriction maps and sites (ie., open reading frames) to the underlying sequence. CAD Gene has a greater range of input methods than Gene Construction kit. Sequences can be loaded which are in native program format, read in with a gel digitizer, or in other formats like GenBank,Staden, and NBRF. The program can "speak" the sequence during entry. This is a distinct improvment over GCK. The program can also write output in GenBank format. If a sequence is in a form that can't be directly read (like part of a word processing document), it can be loaded as a "text" file. The program also allows editing of restriction enzyme data and genetic code data. Unlike Gene Construction Kit, program directly reads GenBank files and their feature tables. This means the program will automatically note the positions of reported genes, promoters, binding sites, etc. Only the old keywords are supported in GenBank features. The program warns you when it finds a feature type it doesn't recognize. In the manual, the authors of the program explain why the new keywords aren't read (they don't seem like good reasons to me). It you have an unsupported feature in your GenBank file, you can always load the file as text, and add it to the sequence manually. Since sequences may have hundreds of features specified in their files, CAD Gene makes some of them invisible by default. You can make any combination visible, however. The interface of CAD Gene is significantly different from GCK. The basic idea of having windows with either text or graphic windows is maintained. Different "views" of the same sequence with different information annotated appear in different windows, as does the restriction map. One feature I liked in CAD Gene was the ability to graphically rotate circular plasmid maps. The program maintains a list of all the sites in a poppup menu at the bottom of the menu. Selecting one makes all those features visible. Some basic MacDraw-type tools are also available, though lines with arrows should be available. The user selects sites in a more MacDraw-type fashion than GCK. In general, I found it easier to learn to manipulate constructs with this program. GCK is more flexible in what it allows, but is harder to use. For example, it is harder to record data about a site in GCK, but more information can be stored per site than in CAD Gene. On the other hand, it was easier to change the style, color, pattern, etc. of a subsequence in GCK. In CAD Gene, I found the option for changing the width of the construct particularly confusing. "Hacking" the program, I accidentally created several useless views of the sequence before I figured out what was going on. Neither program, of course can draw what molecular biologists actually expect to see in a program of this sort. That is, it should be possible to "drag" a defined feature off the main sequence, and have it connected to the main sequence like dotted lines. This is a standard method for drawing vectors and other constructs, and it is puzzling that the programmers have ignored this. Also, one expects to double-click on a feature in the graphic view and have its sequence and/or data associated with it appear in a dialog (like the old Cricket Draw program). To support my point, I let several molecular biologists try these programs. They ALWAYS try to drag the feature/subsequence off the main sequence, and ALWAYS double-click to find out more about a feature! Who's right, them or the programmers? Another unique feature of CAD Gene is sophisticated printing control. At several points in the program, it is possible to pull up a print option dialog. The manual provides instructions for using the print options, which are considerable. The main feature of CAD Gene, like Gene Construction Kit, is support for cloning. CAD Gene has a straightforward approach to this, devoting a specific menu to it with options like restrict, extract restriction fragment, ligate, and edit ends. The program can save blunt/sticky overhang information for any particular site, and knows about compatible ends in ligation. This type of support is necessary for complex cloning projects. A Genealogy database keeps track of the history of a sequence construct, also essential for serious vector design. I did not test this option, since I didn't have the database. The most interesting feature of this package is the wealth of analyses devoted to restriction digests. When a restriction digest is complete, the program shows a series of maps aligned with a schematic of the sequence. The schematic shows features like genes and promoters, and a cursor helps compare restriction site positions to feature positions. It is possible to sort the results of multiple digests in various ways. individual restriction sites may be copied from the restriction site window to the window showing construct features. I find it very easy using this to create a picture of a construct showing only the relevant sites. The same operation is more difficult in GCK, since its restriction analysis is less sophisticated. To summarize, this program is a serious competitor to the Gene Construction Kit. While it does not have the same flexibility, it offers the necessary tools for vector design, and does so in a relatively easy to use way. It supports more methods for entering sequences, and offers good printer support. Since a demo version of the programs is available, I would suggest that anyone contemplating buying a program of this type get demos for GCK and CAD Gene, and compare them for themselves. ************************************************************************** Additional Programs for Sequence Analysis In addition to a general-purpose sequence analysis package, you may want to get some special-purpose programs as well. The following list several programs of this sort. 6)DNA/RNA Builder and Protein Predictor $299, Atlantic Software, P.O. Box 299, Wenham, Mass. 01984, (508)-922-4352 These programs generate graphical analysis of protein structures. Program allows graphical output of peptide hyrdopathy or peptide secondary structure predictions. It is possible to define your own set of secondary structure or hydropathy parameters for analysis. The unique thing about this program is that it uses parameters generated from a "neural net" simulation to do its secondary structure predictions. This company also produces a program for modeling organic molecules, and a program for modeling the 3-dimensional structure of DNA and RNA. The program is written in MS Basic, which makes it really slow. A far more serious problem is that the mac interface was COMPLETELY ignored by the programmer. There is a single window which cannot be moved or resized. This makes it nearly impossible to work with the program under MultiFinder. There are several points in setting up the search where it is possible to "get stuck" and not be able to do anything except quit the program. Overall, the program shows bad interface design which is not excusable in a commercial program. The DNA modeling program has the same interface problems. One good thing about the programs is that graphic output may be copied to the ClipBoard. This means that you can, for example, make a 3D model of a DNA sequence, copy it under the Edit Menu, and paste into a MacDraw document. If you need pictures of a DNA sequence from different perspectives, this is a low-cost solution compared to MacImdat. 7) MacMimic Instar Software AB Research Park IDEON S-22370 Lund, Sweden (note that the following came from an e-mail announcement by Anders Sundin) MacMimic is an application for the display, construction and comparison of molecular models in full 3-D. It runs on the Macintosh II family of computers with a 256 colour monitor and 2 MBytes of memory. A demo of MacMimic version 1.0 is available via anonymous ftp from pollux.lu.se as macmimic_demo.sit.hqx. The Macintosh interface is fully implemented. MacMimic can handle multiple windows, and runs both under the Finder and the Multifinder. Most tools of MacMimic are placed in palettes and are therefore convenient to use. The standard Apple, File and Edit menus are available. Undo works on every command that changes the structure of a model. Copy and Paste can be used to move models between display windows or to the Scrapbook. A picture of a model can be copied from a display window to be included in a word processing document. Printing in black and white or in colour is available directly from MacMimic. Colour printing can be performed on any colour printer with a Macintosh compatible printer-driver. The colours used in printing can be customized through the Preferences dialog. MacMimic can display and handle structures with up to 32000 atoms. A large number of molecules can be displayed simultaneously in each of the windows of MacMimic. Molecules are displayed as stick or ball-and-stick models, each with toggable attributes such as atom labels, atom index and stereo. Colour is used to indicate atom type, but individual atoms or whole structures can be painted in other colours. The 3-D impression of models is improved by a configurable depth cueing. In addition to documents created by MacMimic, structures can be input from the MolLib library of structures, or obtained from a database as a PDB file (Brookhaven Protein Database file) or as an XR file (Cambridge Crystallographic Database file). Structures can also be imported and exported to the Macintosh applications Ball&Stick and Chem3D. MacMimic has a powerful structure editor which can create new and modify existing models. Atom type, charge and hybridization can be set for any atom. The bond lengths and the number of ligands are automatically adjusted. A set of substituents is available from a palette to make instant substitutions of any monovalent atom. The new dihedral angles and bonds are adjusted to make a low energy conformer. Any two models can be joined (substituted) or fused to create spiro compounds or complex ring systems. Chiral control is automatically achieved as all models are displayed and handled in full 3-D. The Enantiomer menu command changes a selected model to its mirror image. The Swap Substituents tool can be used to let any two substituents of an atom switch places. This tool provides a convenient way to switch absolute configurations (R<=>S), equatorial and axial substituents, and cis and trans isomers (Z<=>E). Pseudo-atoms can be placed at the geometrical center of a number of atoms, or at the extension of a bond. Models can be compared by superimposition. Up to eight models can be superimposed (RMS fitted) simultaneously. The MacMimic package includes the molecular mechanics program MM2(87). The MM2(87) User's Manual provides a detailed and authorative description of MM2(87). All computational options (for instance dihedral drivers, restricted motion, etc.) of MM2(87) can be accessed through an interactive input interface. MM2(87) can be run as a background task. 8) DNA Parrot Gel Digitizer System $700, TNT Research, 44 George St. Etbicoke, Ontario, CANADA M8V2S2 416-252-4789 (Voice recognition system about $800 extra for hardware and software) System consists of ÒDNA ParrotÓ software, (now a desk accessory), and a hardware gel digitizer box and mouse. If you are willing to spend some money, you can also get voice recognition hardware, which will allow you to read off the gel without typing to your mac. The box plugs into the serial port of the Mac, and a cable connects the 5-button mouse. The user positions the mouse over the gel band, and clicks the keys corresponding to the base (it ispossible to enter ambiguous bases). The keys light up when clickes,and the Mac says the base name in a male and female voice, flashing the key lights. If you have the voice recognition, you simply say the bases as you read them. A simple solution to sequence entry that works well. It is much less expensive than ÒsonicÓ digitizers, which can run $3000-$4000(!) The new DA version is MUCH faster and easier to use than the SuperCard version of the program. It also echos back to the reader, flashing the appropriate light. You can set it so it only flashes the lights on the input box while reading back the sequence. This is quite useful if you don't want to disturb others with the computer talking! When the voice recognition system is used, the user uses the 5-button mouse as a pointing device, and speaks into a headset. The mac echos back the sound, and flashes the appropriate light on the mouse. According to the manufacturer, the four bases are best entered by reading 'Alfa' for A, 'Charlie' for C, 'Golf' for G and 'Tango' for T. Each letter is trained by the user 3 times the first time it is used, and all the degenerate bases and other commands in DNA Parrot (such as repeat, Select all, Read selection, etc.) can be voice activated. The manufacturer suggests that the system could also be used to read in published sequence data, as well as read in a sequence for a DNA synthesizer. The voice recognition system may be used with other programs besides DNA Parrot, which is another reason to get it. You can make any program and its menu and dialog items voice activated. In summary, the new version of DNA Parrot has become quite a useful product. The desk accessory form is a natural, and should make sequence entry significantly easier. If you can afford voice recognition, this may be very useful. I think the author should consider creating programs to control DNA synthesizers using the voice recognition system. Reading this, some of you are probably wondering why a computer can't be designed to read a gel directly with some sort of video camera. This turns out to be nearly impossible. It is very difficult to program a computer with the knowledge that lets you decide between a faint band and a smear on the autoradiogram, for example. In the future, the most likely solution will be to run gels where the DNA bands are labeled with a fluorescent dye, and their elution is recorded by a detector, which then feeds the information to a computer. Systems of this sort are in use today, but it will be a while till they become universal, due to their enormous cost. 9) Plasmid ARTIST $395 plus 7% sales tax, $5 shipping GeneSystems Computer Software, 755B Loma Verde Avenue, Palo Alto CA 94303 415-493-0972. Commercial Plasmid drawing program, MacDraw-like interface. Allows several text styles, custom arrows. Appears to call PostScript directly, which allows smoother shading and curves than from programs making QuickDraw pictures. Also appears to have some page layout capabilities like desktop publishing programs. Somewhat expensive for a single-purpose program of this type, since the freeware program MacPlasMap provides much of the same functionality (see below). 10) Ball and Stick: Current price US$ 299.- for MacPlus and higher version US$ 359.- for MacII, SE&30 Version University Discount 40% Note: this is an e-mail message by Alexander Falk c/o Inst. f. Chemie Altenbergerstr. 69 A-4040 LINZ AUSTRIA Ball and Stick is a 3D molecular graphics package for Apple Macintosh computers. Its mouse driven menu and dialog user interface is easy to learn and use. It integrates fully into the Macintosh standard environment. Molecular Biologists might be interested in a beta Version available together with the Standard version, which can read pdb files and supports their residue naming. biology, crystallography, and related fields. The scope of the program has now been extended far beyond the needs of mere desktop-publishing by incorporating facilities to examine and change structural parameters via a unique, floating geometry information and manipulation window (/^Info Window^>) and popup dialogs. B&S can display wire-frame, stickball and space filling models of molecular structures consisting of up to 32,000 atoms depending on available memory size. Stereo images, perspective, zooming, and arbitrary orientation in space can be chosen by simple pull down menu commands and user High quality printouts may be obtained taking full advantage of the maximum resolution offered by the printer in use. B&S optionally uses PostScript( ^gray scales on appropriate printers. This applies in particular to the Apple LaserWriter family. B&S is compatible with most color devices currently available (color screens of any size, color printers, and plotters). Color printing is even possible on computers with black and white screens only when an ImageWriter II or LQ are used. Standard TEXT-files (ASCII), provide structure input. Coordinates downloaded from a database query can be immediately read with no or only minor modification. Besides Cartesian coordinates conformational (internal) coordinates (bond lengths, bond angles and dihedral angles) are also accepted as input, which offers a simple way to construct new molecules. Images created by B&S can be saved in the standard PICT-format. This format can be read by many graphic programs available, which may therefore be used to modify the pictures. The Clipboard can also be used to transfer images created by B&S to other applications such as word processing and page layout programs. Animated displays (/computer movies>) can be generated by transferring frames to an animation program like VideoWorks^II or HyperCard . This task is greatly facilitated by two batch operation commands and the HyperCard Stack MovieMaker that comes bundled with Ball^& Stick. = Maximum atom number 8,000 (hardware dependent) 32,000 atoms on a Macintosh II or IIx = 4 model types: - Wire frame - Ball and stick - Simple space filling - Space filling = Depth shading (wire frame model) = Mouse-selectable atoms = Cumulative rotation - of all atoms or of fragments around a coordinate or atom defined axes - via Euler angles = Mirror image = Stereo views = Orientation of planes or axes parallel to coordinate planes or axes = Unique floating information display window for atom specs, distance, bond angles and dihedral angles = Interactive manipulation of geometry parameters = Saving of coordinates and pictures = Batch operation to create animation file series = HyperCard stack MovieMaker bundled with Ball & Stick, used to assemble and play animated displays =99 customizable atom types (radii, patterns, colors) = Customizable program settings = Loading and saving of settings = Fullscreen display and color background option, e.g. for taking screenshots with a camera = 64 -step grayscale on printers supporting PostScript( = Color shading or optional dithering = Faster screen refresh = Complete manual Compatibility: Ball & Stick runs on Macintosh computers from the 512KE upward including the Macintosh II and IIx (at least 1MB of main memory is recommended for good performance). It supports color with appropriate hardware is present (color screens and printers) and is MultiFinder and A/UX compatible. A special fast version for Macintosh computers with MC68020 or MC68030 processors and floating point coprocessor is also available: Ball & Stick II. 11) MacImdat $2950 (50% academic discount) Molecular Applications Group,880 Lathrop Drive Stanford, CA 94305 FAX (415)-857-1754 MacImdat is capable of displaying three-dimenstional structures of molecules ranging from small organic molecules to large proteins. It is capable of several display modes, including "ball and stick", dot surfaces, and space-filling display. The programs is unique for mac modeling in that it can read Brookhaven coordinates for protein structures. Extremely powerful, comparable to protein modeling programs available on graphics workstations like Iris and Evans and Sutherland. But you can't replace an amino acid and do energy minimization like you can in UNIX programs like FRODO. Upgrades and online support available for an additional $500/yr. Nothing in its class on the mac. A must buy if you do protein modeling. To make the program run 3-5 times faster (read:not frustratingly slow)you might want to get a speed upgrade for your mac II to a 40mHz Mac f/x(about $2000, from Apple). The program will be of interest mostly to those working with protein structure/function. Unlike the other 3D modeling programs on the mac, this program has several algorithms specific to display of proteins. The graphics are every bit as good as you would see on a workstation, even if the redraws are somewhat slower. The lack of editing abilities is the program's main problem. Almost everyone who is working with protein is doing site-directed mutagenesis, which means the ability to replace one amino acid with another in the chain is crucial. Without this ability, you are reduced to planning potential substitutions on the basis of an existing structure. The user interface of this program is one of the WORST I have ever seen. Why is it that programmers porting a program over from another system assume we want to go back to command line interaction? In MacImdad, you can't load save, or print anything by selecting a menu option. It is necessary to learn a command sequence in order to do this. It is even more remarkable in that the program runs so well on the mac. Surely if it is possible to make this program run so well, it wouldn't be much harder to make the program at least copy the screen image to the ClipBoard? Or print using the standard print drivers? Or save a PICT file using a save command under the File menu, instead of typing parameters into the command line? This program (plus one year of support) exceeds the cost on a new Mac Si with a big hard disk and extra memory! For that much money we should get the rudiments of the user interface. In closing, if you want to do protein modeling on your mac, this is the best program for it. But don't feel you have to accept the interface--no program for the mac has ever succeeded which did not adopt the standard mac interface (look at TExT or InterLeaf Publisher for examples). Make SURE the company knows you expect more of them. After all, the first version of IBI's sequence analysis program on the mac had similar problems. Expected or Rumored Programs for 1991 DNASTAR-Mac - DNASTAR-Mac - Now called Lasergene, is being released as a series of modules. A shotgun alignment program is already available. Additional modules for keyword,pattern and homology searches of GenBank/NBRF/PIR databases will be released, costing between $500 and $3000. Programs for alignment, protein secondary structure prediction, and DNA/RNA secondary structure will be available in the fall. Presumably like the PC version, the databases will be available on CD-ROM. STADEN-Mac - The mac version will sport a new "common interface" for all versions of STADEN on microcomputers and mainframes. On character-based computers, options will be selected via keys, while the same commands will be displayed in a pulldown menu on macs and PCs running Windows. The cross-system portability is a good idea, but STADEN has never been easy to use, and likely will remain this way, at least compared to 'native' mac programs like MacVector or GeneWorks. It will have an excellent system for sequence alignment from cloning projects. Said to be reasonably priced (lower than $2000?). F. Noncommercial and Public Domain Programs Noncommercial programs are those that are sold informally, by the individuals that developed them. Though they still cost money, it is usually about 10% of a program like MacVector. In many cases, they are a good "first buy" for the novice user. These days (1991), there are a lot of public domain offerings for the mac. Public domain means the program may be distributed freely, as long as it is not sold. The volume of PD software is growing steadily, as developers finally tackle making mac versions of their programs. In using these programs, the general rule of thumb is: If you are a seasoned mac user who is also familar with file formats, interprogram file transfer, etc., a set of public domain programs may be all you need. If, however, you are new to the mac, the integration and consistency (as well as freedom from bugs) of commercial packages will be your best bet. 1)DNA Strider, Christian Marck, Service de Biochimie, Batiment 142, Departement de Biologie, Centre d'Etudes Nucleaires de Saclay, 91191 Gif-sur-Yvette Cedex, France. Old version free, new version about $240 US. Reference Nucleic Acids Research, vol. 16(5), March 1988. Newest version is 1.1. This program has set a standard for how to do sequence analysis on the mac. The author created numerous innovations, many of which have been adapted in the good commercial programs. The ability to create an annotated sequence, help windows, the form of the editor window, have all been copied by GeneWorks and MacVector, and rightly so. Anyone who plans a sequence analysis program has to do better than DNA Strider (DNASTAR, are you listening?) There are two versions of DNA Strider floating around. The earlier 1.0 version was distributed freely. The new version costs about $200 US dollars, and may be ordered directly from Christian Marack at the above address. Its main new features allow saving the results of an analysis as text or PICT files, instead of simply printing the topmost window. Further upgrades are expected in 1991. If you are using the old version, you should get the new one to get on the upgrade list. Marck was the first programmer to use the full power of the Mac interface to simplify the analysis of DNA and protein sequences. The various options in the program are controlled by Hierarchical Menus, making it much easier to choose among multiple options. I will describe each menu in turn: The File Menu: DNA Strider distinguishes between different sequence files. Under the Menu Option NEW, there are suboptions to make a DNA, RNA, protein, or DNA degenerate sequence. The Open option allows you to show only the above file types in the file list. Several Finder functions are also provided, which are extremely useful such as Save a Copy, Duplicate, Make new Folder, and Delete a File. These functions are useful when you don't have enough memory to run DNA Strider under MultiFinder. When a sequence is loaded, it appears in an excellent sequence editing window, which was clearly the model for the editing windows in MacVector and GeneWorks. It looks a little like the text editing window in the DNA Inspector, but is much more sophisticated. The sequence is displayed along with comments and numbering. When you enter or remove sequence areas, the numbering instantly updates. Positions of the cursor are given as absolute position (eg., 500), and relative (eg., 444 from end). The frame of the cursor (1,2,or 3) is also given! This kind of information is necessary for intelligent sequence editing! The edit menu incorporates the standard Cut/Copy/Paste of most Mac programs, and many additional functions besides. These functions apply to the sequence editing window, and you cannot use Copy to save a graphics image to the ClipBoard with the 1.0 (free) version. The 1.1 version allows printing/saving/copying to ClipBoard of any window the program creates. The Select option has several suboptions; Select All, Select From......To, from beginning of sequence to cursor, and from cursor to end of sequence. These options are very useful in copying restriction fragments, etc. from a larger sequence. The only thing missing is the "Show ClipBoard" option. This Find menu allows you to move the cursor and selection range. One can jump to the the beginning of a sequence, its end, a position within the sequence, or to a particular subsequence like a restriction site. An especially useful feature is the Find Next open reading frame (ORF). This allows you to select the next open reading frame from your position. The ORF sequence may then be saved using the Copy command in the Edit Menu, and put into another editing window. You can also translate it,and a new protein window is created to hold the translation. It amazes me that this feature never existed in the other programs available for the Mac. This Convert menu contains options for changing the type and format of the sequence in the editor. It works by the user selecting a region of a sequence, and then selecting one of the Conv options. As an example, If the Next ORF was selected, selecting 5'-3' protein in Conv creates a new editing window with the translation of the sequence in it! Instant, intuitive translation!!!! One can also convert a sequence to 3'-5' protein, RNA, and DNA degenerate. The second set of convert options affect the topology of the sequence. One can reset the origin +1 position of the sequence, circularize/linearize it, and generate the invert, antiparallel, and compliment forms of the sequence! These options only work on a selected portion of the sequence. An example of the utility would be to 1) select a sequence region corresponding to a restriction fragment, and then 2) select compliment from the Conv Menu. This would "flip" the sequence, exactly as if the sequence was restricted out and religated in. This is another extremely useful function for those doing site-directed mutagenesis. A final option allows you to keep a little window around with the DNA degenerate alphabet, for reference. The final few menus do familiar functions such as restriction, translation, codon usage, and so on. The basic utilities are present, but there are as yet no routines for comparing sequences to each other and aligning multiple sequences. You do get: 1. Restriction maps (circular and linear), several options such as all unique sites. The program draws very nice maps. One limitation is that the user cannot specify subsets of restriction sites to be grawn in the graphical map, only unique sites. 2. Editing the Restriction enzyme list 3. Translation, several sub-options. 4. Codon usage, using several genetic codes, and amino acid usage for proteins. 5. Digests with one or more enzymes. 6. A window listing enzyme data 7. Acidic/Basic map of a protein. 8. Hydrophylicity, Hopp-Woods, Kyte-Doolittle, and others. Draws nice plots, but rather small. It would be nice to resize these windows. 9. Hydrophobic Moment of a protein. 10. Data windows for the genetic code, and amino acid data. This list is far smaller for the big commercial programs. If you need the extra functionality of these program, you will have to buy them instead of Strider. For example, the output graphics are not linked in any way to the original sequence, unlike GeneWorks. Unlike the commercial programs, you can't "zoom" in on an analysis, though the newer version of the program allows you to resize graphic output. The program has no provisions for gel digitizer entry of data, or speaking a sequence, which is standard in the commercial packages. Similarly, there is no ability to search the GenBank/EMBL files for homology, or dot-matrix alignments. The program provides only a couple of protein analysis routines, in contrast to the dozens in the commercial packages. Finally, the program is ignorant of feature tables. This is a serious problem for those working with sequences from the databank in vector construction. Hopefully, now that the new GenBank feature table format is out, the author will include this feature. The last menu lists the windows currently on the screen. Selecting a menu item brings the window to the top. This feature is also found in the commercial packages. In summary, DNA Strider is an fantastic program. I advise anyone who is going to do sequence analysis on the mac to get it. It may be all you need! It is a good starter program for beginners, because it is easy to use. It will teach users what a computer can and can't do for them, and help them make a decision on a commercial package later. Support the author and get this program! 1a)Gene Jockey, BIOSOFT, 22 Hills Road, Cambridge CG2 1JP, UK ~ $500 US Commercial program comparable in features and power to DNA Strider, and shows the same intelligent interface design. A demo version of the program is available from the company. GeneJockey has a reasonable sequence editor. I like Strider better, since it lists the the starting and end nucleotide for a mouse selection. GeneJockey reads its own native format as well as MacVector and DNA Inspector files. Interestingly, it does not read GenBank files, though it can format a sequence as GenBank, UWGCG, NBRF, and GenBank. Since the database search module uses GenBank files, it is strange that they cannot be input. You can open a GenBank file directly only by purchasing the entire GenBank library, and using the index files to pick a sequence. If you try to load a GenBank file directly (or any other file whose format GeneJockey doesn't recognize), it appears in a text window. You can then copy and paste the sequence into an editory window. The default text window is an excellent idea, since it allows you to open almost any file your sequence is in, and select it out of the comments, text, etc. As mentioned above, the Format command allows the user to write an editor sequence to GCG,Staden, MOLGEN, GenBank, and other formats. Another useful function is the interleaved command. If the sequence window contains a selected open reading frame, it is possible to make output with the translation listed only under the open reading frame. The function,is often needed but seldom implemented by sequence analysis programs. GeneJockey supports multiple genetic codes, and has the ability to edit them. Only the new version of DNA Strider supports this, to a lesser extent. You can also save pictures created by GeneJockey as PICT files. The most important feature of GeneJockey, in my opinion, is the alignment function. Though it falls short of the alignment routines found in programs like Staden or UWGCG, it nevertheless should be enough for relatively small cloning projects. Two sequences at a time can be compared, and combined into a contig. You can also do alignment of two protein sequences in the editor, making a contig, but GeneJockey does not support aligning several DNA or protein sequences at once. To aid in alignment, it is possible to do dot-matrix homology. The program can search sequence and/or comments of a DNA sequence database, (wildcard searches possible). The interface for search is much easier to understand than GeneWorks, of MacVector, but offers less flexibility. Biosoft does not provide GenBank sequences to use in the search. You will have to get them from IntelliGenetics (see elsewhere in this document). GeneJockey has no protein analysis functions except for hydropathy plot. In this respect it is inferior to all other analysis programs except DNA Inspector. In contrast, the restriction analysis module is excellent. A useful feature shows the actual cut site in the sequence with staggered ends, simply by clicking on a restriction site in the graphic. Another interesting feature of GeneJockey is the Communications window. It provides simple terminal emulation sufficient to logon to your local VAX or mainframe. It is possible to logon to a remote computer, retrieve a sequence, and copy/paste it into an editor window. This was a feature that existed in the old MacGene, and was apparently planned for DNA Strider but never implemented. It may make GeneJockey an attractive solution for those who are already using a mainframe program, but want to transfer some functions (such as sequence editing) to the macintosh. The documentation for GeneJockey is excellent. It offers a series of tutorials, each concentrating on a specific application of the program. The manual also covers subjects that are usually left out of microcomputer documentation, but are important nontheless. For example, the session on the terminal window describes an actual online session with a mainframe computer. The section of file formats covers how to tranfer files from an IBM PC to the macintosh. This is very useful information that is hard for novices to find out elsewhere. To summarize, GeneJockey is an excellent first program for its price. It supports the "classic" macs, does almost everything DNA Strider does, has additional features like sequence alignment and database searching. Using the communication window to access mainframe analysis will fill in any gaps in its analysis routines. It is a superior choice to DNA Inspector, which sells for about the same price. 2) Mac Sequence analysis programs, Bernard Bellon, Laboratoire de Genetique et Biologie Cellulaires--CNRS, Faculte de Lumminy, Case 907, 13288 Marseille Cedex 9, France. Programs cost $180 for first version, $270 for program plus two upgrades. Another native mac comprehensive sequence analysis package. Designed to run on all Macs. Appears strong in the ability to save results directly into a editable scrolling window on the screen, or graphics window. Speech functions. Dot matrix has 'point and click' to check alignment on graph at sequence level. 3)Public Domain Programs for Molecular Biology I have ported several Basic and C programs from IBM PCs and mainframe UNIX machines to the Macintosh. You can get more information about them by writing to the following address: Dr. Peter Markiewicz Dept. Microbiology, MBI (UCLA) 405 Hilgard Ave. Los Angeles, CA 90024 213-825-8924 Bitnet Address: JMILLER%VXBIO.SPAN@STAR.STANFORD.EDU Note that most of these programs are also available from the various servers; U Houston, IuBio, EMBL, etc. The following Programs are currently available: (A) DNANALYZE--general DNA sequence analysis program. Runs on old Macs(except 128K), Mac Plus, Mac SE, and Mac II. Doesn't run on the new mac f/x, impractical to make it do so. Similar in power and performance to early versions of The DNA Inspector. Prints to Imagewriter or LaserWriter (the latter only if output is short). Simple graphics can be saved by 'screen dumps' into MacPaint file. Can't print to the LaserWriter if output >1 page. (B) MOLGENJR--general DNA and protein sequence analysis program. Runs on old Macs (except 128K), Mac Plus, Mac SE, and Mac II. Doesn't run on the new Mac f/x, and impractical upgrade it. Similar in power and performance to early versions of The DNA Inspector. Prints to Imagewriter I and II, Epson, and generic dot-matrix printers. Uses the colored ribbon on the ImageWriter II to create colored printouts if hydropathy, Garnier structure predictions, and dot-matrix homology. Can't print to LaserWriter if output >1 page. (C) FASTP-- I no longer distribute this early version of William Pearson's FASTA package. If you want to search DNA or protein sequence databases on your mac, contact him for the program (see address below). (D) THREE--align three protein sequences for maximum homology. This program uses a new algorithm which takes less time and memory than traditional alignments. From Russell Doolittle's lab. (E) t-RNA search package--search a DNA sequence for prokaryotic, eukaryotic,intron, and mitochondrial t-RNA genes. (F) MacDNA--basic sequence analysis program, with PC-style interface,by Robert Scheif. (G) LIGATE AND CIP--programs determine optimum conditions for ligation and dephosphorylation reactions. Ports from PC by Hal Jenson. (H) MacPlasMap, Plasmid Draw, PROPLASM --- plasmid drawing programs. MacPlasMap is comparable to MacDraw in power and ease of use. (See review ,below). (I) Gel Sizer(HyperCard stack)- this stack figures out the sizes of DNA or protein molecules by their gel electrophoresis mobility relative to a standard. (J) PHYLIP phylogeny inference package for the Macintosh-- this large collection of programs allows the construction of phylogenetic 'trees' from sequence and taxonomic data. Interface is un-maclike, but programs run well. You need to be a specialist in evolutionary biology to create input and interpret the output. (K) Enzyme Kinetics/Scatchard Plot program/Endocytosis analysis program. Of use to those doing binding assays and enzyme kinetics. (L) PUPPY--Graphical representation of DNA sequences. (M) Multiple sequence alignment programs (little macs) programs AlignSeq, GenAlign, Ralign, DottyPlot, and THREE. All these programs provide for aligning one or more DNA or protein sequences on Mac Plus/SE level systems. (N) Multiple Sequence Alignment (Mac II) Vax ports of Doolittle's protein sequence alignment package, CLUSTAL, and others. Require Mac II with 5 or more megabytes of RAM. (0) PLOT/A protein analysis package A steadily growing set of programs for analyzing protein sequence. Currently, the package consists of 10 programs (compiled version) or 14 (interpreted versions). A common user interface throughout allows the user to create plots of a protein sequence's hydrophobicity,chain flexibility,secondary structure, and search for motifs like membrane-spanning peptides, leader sequences,and more. The plots created may be printed or exported to MacDraw or a MacPaint-type program. (see review below). P) MacPatternSearch Created by Rainer Fuchs at EMBL, this program can search the PROSITE protein motif database to find matches to your sequence. A very well-written mac-style program. You can also get it directly from the EMBL server (see end of this article) Q) MacTargSearch A little clumsier and slower than MacPatternSearch, MacTargSearch was developed by Jim Goodrich to search DNA sequences for promoter sequences. It can find any complex consensus sequence, with "spacer" regions between them (like those between the -35 and -10 regions of an E. coli RNAP promoter. R)MFOLD and LoopViewer Programs developed by Don Gilbert at IuBio, they implement Zucker's algorithm for folding RNA molecules into energy-minimized secondary structures. LoopViewer can take the numerical output of MFOLD and convert it to a MacDraw-type picture showing the hairpins. Very interesting and powerful programs. S)HyperCard Stacks for Molecular Biology/Biochemistry Lots of HyperCard Stacks, too numerous to name here. T)MacLigand and Endocytosis Programs for computing scatchard binding, and endocytosis calculations. Please contact me for more information, and to find out how many floppies you need to send to get the programs. Several other programs will be available in the future: A) More protein sequence alignment programs. I am porting several programs for protein sequence alignment described in the March, 1988 Nucleic Acids Research computer issue.  4)Genetic Linkage analysis/Pedigree analysis programs. These are programs of interest to those in human genetics. I have the VAX version of LIPED. LINKAGE is also being ported to the mac. You can inquire as to the status of LINKAGE by contacting: Jung Ott College of Physicians & Surgeons Columbia University, Dept. of Psychiatry, Box 58 722 West 168 Street New York, NY 10032 212-960-2504 FAX: 212-795-5886 EARN:BITNET: OTT@NYSPI You can get a pedigree drawing program, Pedigree Draw, from the following address: Pedigree/Draw $30 Department of Genetics Southwest Foundation for Biomedical Research P.O. Box 28147 San Antonio, TX 78284 U.S.A. 512-674-1410 Comments or questions to Paul Mamelka at: "paul@darwin.sfbr.org" Pedigree/Draw comprises 2 programs that allow creation,editing, and drawing of genealogical diagrams. Programs use mac interface, with interactive editor for entry, updating, and location of pedigree records. Note that you can get a Mac version of the HGML library from the Yale Dept of Genetics, at the following address, for free: Yale HHMI Human Gene Mapping Library 25 Science Part New Haven, CT 06511 BITNET:GENEMAP@YALEVM FAX:(203)-786-5534 HMGL is the mac version of the human genetic map database, with one file for each chromosome. It is well done, compiled by nomanclature,chromosome, and DNA committes of the Human Gene Mapping Workshops coresponding to the tables as printed in the HGM10 Publication:Volume 51, Numbers 1-4 of Cytogenetics and Cell Genetics(1989) and available free from the Human Gene Mapping Library. Data regarding probes and polymorphisms on the diskettes, as compiled by the DNA Committee, are also contained in the PROBE and RFLP components of the HGML Database. Several fields on the diskette contain pointers to specific records in these components, as described below. References to published literature on the diskettes consist of a five character LIT Ref file number, which point to one of the 13,000 literature citations catalouged in the LIT component of the HGML database. For comments, questions, or copies of the diskettes in other formats, contact Dr. Kenneth K. Kitt at the above address, or at Dept. of Human Genetics Yale School of Medicine 333 Cedar St. New Haven, CT 06520 BITNET:KIDD@YALEMED FAX:(203)785-6568 4) Mac Plasmap, v. 1.82, by Jingdong Liu, Dept. Biology, University of Utah, Salt Lake City, UTAH, 84112 801-581-6307 $20 EXCELLENT program for creating plasmid maps. Allows MacDraw-like interaction to create restriction sites, gene regions, polylinker regions, etc., on circular plasmids.Some features handled through dialog windows may be done entirely by mouse in later versions. Excellent LaserWriter output, and the Copy command puts the complete map into the ClipBoard for use in other "Draw" programs. It also can save a MacDraw PICT file of the plasmid picture directly. Though it has much in common with MacDraw, you can't resize and drag parts of the map around by using the mouse. Instead, you go through dialog boxes where you type in numerical information. I have found that people who don't use the mac a great deal tend to be confused by this approach, and often repeatedly make the same mistake repeatedly in the dialog. The user must know numerical positions of restriction, polylinker sites, as there is no function in the program which will find restriction sites in a sequence. Only the picture and associated comments is present in the output file, rather than the DNA sequence itself. It is the best noncommercial plasmid drawing program. It lacks several features found in programs like Plasmid Artist, and is a little harder to use. However, it gets the job done, and there is nothing else out there in freeware/shareware that is comparable. The author requests a puny $20 for his trouble. Aw, come on now. You just spent >$5000 for your computer, and you don't have money for this? 6) PLOTA protein analysis package MS Basic interpreted versions: Dr. Angela Luttke, Arnimstr.2 5000 Koln 30, GERMANY Compiled versions: Dr. Peter Markiewicz, see address above Programs are also available on the EMBL server as .hqx-encoded files. A steadily growing set of programs for analyzing protein sequence. Currently, the package consists of 10 programs (compiled version) or 14 (interpreted versions). A common user interface throughout allows the user to create plots of a protein sequence's hydrophobicity,chain flexibility,secondary structure, and search for motifs like membrane-spanning peptides, leader sequences,and more. The plots created may be printed or exported to MacDraw or a MacPaint-type program. The compiled versions have several improvements over the interpreted versions, including the ability to directly read files created with DNA Strider,GenBank,Pearson format, and more. For this reason,only the compiled versions are discussed here. When you start one of the PLOTA programs, three menus and a dialog window appear. The dialog window has buttons to load sequence file(s), exit the program, to display a help message about the program, and to specify the subsequence you want to analyze in the protein sequence. If you click the load button, you can load files in several formats, including plain sequence, GenBank, EMBL, BioNet, Pearson, DNA Strider text files, and DNA Strider PROT files saved in the normal way. The ability to read DNA Strider files is a great help. This way, you can directly analyze a sequence further after entering it in the DNA Strider editor, without having to create a separate text file. When the file is loaded, the programs displays its length in the "Residues" part of the window. Edit fields are provided for selecting a subsequence of this sequence to analyze. This feature is important, because some of the programs can't analyze sequences greater than 500 amino acids in length. This is a limit imposed by the way the Mac copies draw pictures to the ClipBoard, rather than a programming limit. Once the file is loaded and parameters are set, clicking OK brings up the graph window. The graph is drawn as you watch, and the scale can be specified in most of the programs. At the same time, the programs can write a text file containing the actual numbers used to create the graph. This output can be loaded into a spreadsheet/statistics/plot program for further analysis. One of the best features of the PLOTA programs is seen when the graph is drawn. The user can click on the graph, and the program will write the sequence position. For example, if a strong hydrophobic region is seen in the plot, clicking on either side of it writes the sequence positions. This allows the subsequence to be analyzed in greater detail by a program looking for alpha helices, or a helical wheel projector. The PLOTA programs have three (unfortunately) nonstandard menus. The File menu allows the user to copy the plot to the ClipBoard, print the plot, or quit. The second Program menu allows the user to rapidly chain to another program in the set, without having to quit the current program and start the new one. If is a time saver if you don't use MultiFinder. If the Mac can't find the program, it prompts you to find it. Afterwards, it "remembers" how to find that program. This way, you can put the programs wherever you want, and they will be found after an initial "learning" period. The last Options menu allows the user to set the color of the plot, and the output file type. The colors selected will be echoed to any color printer (such as a HP-PaintJet,or color LaserPrinter), and the colors can be set even you are using a black and white mac! The colors will appear in any program you paste the plot into (such as MacDraw) if it supports color. The output file type allows the user to specify the file "creator" of the text output. In simple terms, this means that you can make the output a MacWrite text file so that when you click on the file, MacWrite will start. Both color and file type settings are saved between program runs, so you don't have to set them each time. The biggest limitation of the PLOTA programs is file size. The secondary structure program can only analyze sequences about 500 amino acids in length, and the rest of the programs are limited to about 1000 amino acids(larger sequences can be loaded, and portions analyzed, however). I wish that the File menu was "standard mac"in appearance. Despite these minor problems, the PLOTA programs are excellent for users wishing to analyze proteins. Since more are being written by Dr. Luttke all the time, they should come to form a comprehensive package in the near future. Programs run on all Macs except 128k,not copy protected. It is possible in the future that there will be versions of the PLOTA programs which compute average properties for a set of protein sequences aligned by another program. 7)Pearson Sequence Analysis Programs,William R. Pearson,1611 Westwood Rd., Charlottesville, VA 22901 804-924-2818, William Pearson, creator of FASTP/FASTN/FASTA, has mac versions of these extremely popular programs. They may lack the Macintosh user interface, but they will get the job done.The programs cost about $50 US dollars, but the prices change constantly, I strongly suggest you contact him for pricing information before ordering, as prices may change. (A)General Program Package: Contains FASTA, TFASTA, and others, for searching GenBank DNA or Protein sequence databases for homology. Also multiple alignment algorithms, and other very useful programs. Some programs save their graphics as PICT files, which can be used in MacDraw and SuperPaint. (B)FASTA Sequence Database: Gives you the sequence database on the Mac. Used with the above programs. An excellent investment. 8) PROPLASM, plasmid drawing program, P.F. Lurquin, Program in Genetics and Cell Biology, Washington State University, Pullman, WA 99164-4350. Plasmid drawing program, similar but much less sophisticated than MacPlasMap mentioned above. Lousy mac interface. Has options for saving the plasmid picture in different formats, text and graphics. Not worth it now that you can get MacPlasMap (see above). 9) GenBank Search (HyperCard) available via anonymous FTP to Iubio.Bio.Indiana.Edu, cd [archive.molbio.mac], Get GBSearch.Hqx. This file is a binhex encoding of a stuffit archive of the stack. (review by elliot lear 10/90, on electronic mail) The GenBank Search is a Macintosh Hypercard stack for interacting with the computer services of GenBank, the molecular biology databank. The GenBank computer provides two generally useful services for molecular biology by electronic mail: * Searching of the current GenBank databank for similarities with a query nucleic acid or protien sequence of yours, and * Retreival of a sequence in the current GenBank by locus name or accession number. To use this GenBank Search stack, you must have: -- a Macintosh that is connected to the national Internet network, thru an Appletalk or Ethernet cable, -- your Mac must have the MacTCP communications software, and Hypercard version 1.2 or later, from Apple Computer, -- you must have a valid Internet return mail address. This stack includes a simple sequence editor, so you can enter your sequence directly here, and it includes features to read from disk a variety of sequence formats. The Search card then takes whatever sequence you have in the Editor card, bundles that with some information on how to perform the search, and sends it off thru the Internet network using SMTP (simple mail transfer protocol) methods. For the Fetch card, you need to supply an Accession number or GenBank Locus name, which is send off by SMTP to another maildrop at the GenBank computer. GenBank Search stack is available via anonymous FTP to Iubio.Bio.Indiana.Edu, cd [archive.molbio.mac], Get GBSearch.Hqx. 10) AUTHORIN FREE,IntelliGenetics 700 E. El Camino RealMountain View, CA 94040 (415)-962-7300 This program, available on both Macs and PCs, is designed to standarize the process of submitting a DNA or protein sequence to a sequence database like GenBank, PIR or EMBL. Since most journals now require proof of submission before they will publish a sequence, it is an essential program for any molecular biologist. Also, it doesn't cost a thing! The managers of all the databases have strongly urged us to use AUTHORIN instead of a text form, because the program prechecks entries, making their uploading onto the databases that much faster. The mac version was released in September. It is a true mac-style program, with multiple windows for the different portions of a sequence submission. Such information as relevant journal articles, organism, strain, and FEATURE (yes, they're here too!) tables are handled efficiently within the program. The program contains a sequence editor so you can type the sequence directly into AUTHORIN. Alternatively, you can import a sequence only file from another program. I have worked with the first release of AUTHORIN, and found it to be buggy. In some cases, the program crashed on a Mac IIsi. I hope that IntelliGenetics will take the trouble to fix this problem. I find that the program is easy to use. Otherwise, I have few complaints. IntelliGenetics could have given us a PC port, but instead chose to develop a true mac application. Lets support them by reporting problems so they can be fixed in future releases. I urge everyone who works with DNA or protein sequences to write to IntelliGenetics for this program. G. DNA and Protein Sequence Database Sources: 1)GenBank Sequence database You can now get GenBank on Mac 800K floppies from GENBANK, c/o IntelliGenetics, or on CD-ROM (preferred) GenBank c/o IntelliGenetics 700 E. El Camino Real Mountain View, CA 94040 (415)-962-7300 The price is between $125 and $175, depending what disk format you want and method of payment. Note that this database is provided with MacVector,MacMolly, and GeneWorks, so you don't need it if you get these programs. It can also be ordered for the FASTA package through William Pearson (see above). 2)Prosite Database In: EMBL software database, as StuffIt .hqx file, Postfach 10.2209 D-6900, Heidelburg Federal Republic of Germany Telephone (+49) 6221-387-258 Electronic mail: "NETSERV%EMBL.bitnet" or NetServ@EMBL-Heidelberg.DE (Internet) (send the message "help" to get started) This database contains a series of common protein "motifs", such as binding sites, leucine zippers, zinc fingers, etc. It can be searched using MacPattern, also available from the EMBL software library. H. HyperCard,FaceWare, and Writing Sequence Analysis Routines on the Mac 1)Writing Sequence Analysis Software for the Mac The IBM PC has been called the "programmer's computer" because it is easy to develop programs supporting a command-line or menu tree interface. Conversely, the Mac, while easy to use is hard to program. Many people who have written sequence analysis and alignment routines on other computers have shied away from the Mac, since programming anything like a typical mac interface takes a great deal of learning on the programmer's part. The following describes some useful programmmer's tools for creating a mac-style interface without sweating blood. 2) Hypercard and its Impact on Molecular Biologists HyperBasic: Teknoys, Inc.,3923 Coconut Palm Drive, Suite 111 Tampa, FL 3361 1-800-873-3494 In 1987, Apple started bundling HyperCard with all the Macs it sells. This program has the potential to replace about 40% of commercial software with free and Shareware "stacks". Utility programs like electronic checkbooks, notebooks,address lists, and the like are doomed, because they can be created much more effectively in Hypercard. It is a powerful, extremely simple to use database. Programming using Script is easier than Basic, and allows creation of programs impossible for nonprofessionals to create until now. With the addition of XCMDS, additional features can be added, and if one uses SuperCard or other HyperCard clones, additional features like color and PICT format pictures may be included. What does this mean for us? It will not replace sequence analysis programs entirely, unless someone decides to create one within HyperCard. Expect "data-management" stacks for DNA gel and vector sequences, or management of sequencing projects. For example, the user could draw a picture of the sequene they are working with, and program HyperCard so that clicking somewhere on the picture would bring up a editing window with the sequence from that area! Recently, Russell Doolittle demoed a program of this type in SuperCard which displayed database alignment results from a supercomputer analysis. It should be simple to write stacks for input from any gel digitizer that can be connected to the Mac's serial port. This has already been done with the DNA Parrot program described above. The simplicity of Hypercard means that these types of programs can be developed by molecular biologists, rather than companies. Another area where HyperCard shines is in communication with other computers. For example, the programs Hyper Search and HyperFTP both access unfriendly host mainframes over electronic networks. The user is spared learning the formal commands for doing these tasks. Things have gotten even better with the introduction of HyperBasic. This is a compiler for the Basic language which creates XCMDS and XFCNS for HyperCard. A programmer would use HyperCard stacks to store sequences, and write routines in Basic to analyze them. This combination might prove useful for programmers who want to distribute specialized sequence databases. For example, one could distribute tRNA sequences, and write HyperBasic functions to fold them up into a cloverleaf without leaving the stack. I. FaceWare 1310 N. Broadway Urbana, IL 61801 217-328-5842 This is a set of resources which allow programmers who want to move their algorithms off a VAX,UNIX,or PC environment onto the Mac. The hardest part of Mac programming is creating the typical user interface of windows, menus, buttons,etc. FaceIt, which works with every compiler available for the Mac (including QuickBasic) creates text, graphics, spreadsheet, plotting, and terminal windows with only a few commands. It works especially well with fortran, making it the method of choice for moving fortran programs to the mac. Though there are other "runtime libraries" out there which help create a mac interface, I feel that FaceIt is easiest to use if you aren't a Mac programmer already. The majority of programs that I distribute were ported to the mac using this package. I should note that because FaceIt is so resource-based, you will have to become familar with ResEdit to use the package. The advantage of this is that all your text in menus, windows, etc., is stored in a form such that it can be modified without recompiling the program. This makes it easy to make foreign language versions of a program, important for scientific programs which often have very wide distribution. It is also possible to write FaceIt modules yourself, which will be distributed by the company. Because it is easy to write them, many third party developers have created modules for special functions. The costs of using the FaceIt package are quite low. The basic text editor/spreadsheet/graphics/dialog package costs $75, and the third-party modules for terminal, plotting, and help windows average about $50. Anyone considering writing or porting sequence analysis programs to the Mac should get this package. J) Journals/Magazines/Publications of Interest 1) Computers in Biomedical Research--publishes many algorithms and descriptions of programs for DNA and protein sequence analysis. 2) Computers in Biology and Medicine--tends to be more medical/database oriented in the type of articles, also analysis of medical data. 3) Binary ,Computers in Microbiology--this is a new journal, the others are much more established. 4) BioTechniques, Eaton Publishing Co., 154 E. Central St., Natick, MA 01760 Subscriptions about $95/yr--Has a regular section on computer analysis of biomedical data. Has run reviews of software available for micros in the past. Runs the new online service, BioTechNet, which contains all the mac public domain software I distribute, as well as numerous unpublished articles on methodology. 5) Biotechnology Software, Kevin Ahern editor, c/o Subscription Dept., Mary Ann Liebert, Inc., Publishers, 1651 Third Avenue, New York, NY 10128. Subscription about $90/yr--This journal is the best one for looking at mac sequence analysis software and related topics, as well as PCs. The Jan 1991 issue has extensive reviews of both MacVector and GeneWorks. J. Online Services, Commercial Electronic mail and ftp are rapidly changing the character of analyzing a DNA or protein sequence, or retrieving one from a database like GenBank. Today, it is possible FOR NO CHARGE to send your sequence to the GenBank server, have it rapidly (a few minutes) searched against the GenBank library, and have the results returned in an electronic mail message. It is similarly possible to report a new sequence to GenBank/EMBL in this way, in fact, many journals now require that you do this before publishing your sequence! The same servers you access also have loads of public-domain software for the taking. You get the programs as mail files or through ftp, decode them with commercial or shareware "unstuffing" programs, and presto! You have the programs running on your mac. Okay, mabye its not that easy. Interacting with mainframes is difficult for users friendly with the mac environment. What is necessary to do this is too complex to go into here, but I distribute a HyperCard Stack describing electronic communications for the novice (1991). What follows is for those who have some experience with electronic mail and/or ftp, but haven't begin to exploit the resources available on the biomedical networks: Commercial: 1.BioTechNet: This network is run by BioTechniques magazine, and is frequently advertised there. Access is from most countries in the world. Enrollment consists of a $25 one-time fee, and additional connect charges (no minimum charge). Call (508)-655-8282. Presently, BioTechNet is NOT connected to BitNet or InterNet. BioTechNet contains several bulletin boards, general information, and program archives: Calendar Communications Electronic Mail Conference Center Employment Center Forum Industry News Marketplace Member Directory Network Office Polling Research Library Societies One useful thing about this online service is that many programs described in BioTechniques can be downloaded from the server. It also has hundreds of technique articles which weren't published in BioTechniques, and can ONLY be accessed via the service. Rapidly growing, targeted specifically at biomedicine. America Online: General online service aimed specifically at mac users, features a "point and click" interface that looks exactly like your normal mac. Recently, a Biotechnology section has been started, and it has bulletin boards and software for downloading. The SIG is located in the Special Interest Group Folder of the Macintosh Business Forum (Keyword: MBS). A formal statement from the service follows, since I haven't used it yet. A Science and Engineering Special Interest Group (SIG) is now on America Online. The Group will foster the exchange of ideas and information based upon several goals. The goals are to network scientists and engineers who currently or are interested in using the Macintosh computer in the research environment, address present and future needs and concerns of users relevant to the use of the Macintosh in the laboratory, serve as an educational/informational tool by providing new and relevant ideas applicable to various research environments, and demonstrate the presence of the Macintosh in a large variety of disciplines of science and engineering. These goals are accomplished via 1) a message board which contains message folders of members's interests, discussion of technical applications or problems, 2) a libary which contains public domain/shareware scientific or engineering applications, and 3) live interactive conferences which will cover topics of relevant interest along with guest speakers. K. Noncommercial Academic Services: These are services you automatically have access to, if you local host computer (often a VAX) is hooked up to BitNet or Internet. They are described in more detail in my stack on electronic communications, but here is a brief list, and where to send "help" messages. a) GenBank Sequence Retrieval System: To use the GenBank Retrieval System, send an electronic message to RETRIEVE@GENBANK.BIO.NET containing as text (leave the Subject: line blank) either an accession number, or an entry name, but not both. The message text should contain exactly one word. The data banks are searched in the order: GenBank New Data, GenBank current release, EMBL New Data, EMBL current release, GenPept New Data, GenPept current release, and Swiss-Prot until a match is found. If an entry exists in both GenBank and EMBL with the same accession number (the usual case), a query on the accession number will return the GenBank version of the entry. If the EMBL-format version is required, it can be retrieved from the file server at NETSERV@EMBL.BITNET (for instructions send a message containing the line HELP to that address). b) GenBank Sequence Data Submission Form An electronic version of the sequence data submission form used by the sequence data banks is also available through the RETRIEVE server. To receive a copy, send a message containing the word DATASUB as the only line. Instructions for completing and submitting the form are included. If you have any questions or comments, feel free to mail them to RETRIEVE-REQUEST@GENBANK.BIO.NET. Note that GenBank distributes programs for both macs and PCs, called AUTHORIN, which simplifies the production of the sequence data submission form. I'll review this program in a future version of this newsletter. c) Online FASTA The FASTA program allows you to send a specially formatted mail message containing the nucleic acid or protein query sequence to the FASTA Server at GenBank. A FASTA sequence similarity search is then performed against the specified database using the FASTA program developed by William Pearson and David Lipman as described in their paper: Pearson, W.R. and Lipman, D.J. 1988. Improved Tools for Biological Sequence Comparison. Proc. Natl. Acad. Sci., 85: 2444-2448. Obtaining Help Messages are sent to: "SEARCH@GENBANK.BIO.NET" If you would like to receive instructions on using the FASTA program, send a mail message to the address above containing the word "Help" on a single line of the mail message. Leave the Subject line in the mail header blank. The help text will be updated when new information is available for FASTA searches (such as new databases on-line). For additional help on using FASTA, contact GenBank at (415) 962-7307 or send an electronic mail message to the address: CONSULTANT@GENBANK.BIO.NET d)Getting Mac Programs off The Servers: A) The EMBL Server - contains the same sequences as GenBank, plus a very large collection of mac and other software. Supports getting sequences and programs via e-mail. The address of the EMBL server is: "NETSERV%EMBL.bitnet" You can send the following command through the mail to reveive a list of mac software: GET MAC_SOFTWARE:filename.ext B) The U Houston-GenBank Server - contains the GenBank archives, and archives of mac and other programs. Supports e-mail distribution of programs and sequences. The address of the Houston server is: genbank-server@uhnix2.uh.edu (INTERNET/ARPANET) You should send the word "help" in your subject to get a help file for this server. Subject: send mac help D) The IuBio Server The address of the IuBio server for e-mail is: "GILBERTD@IUBIO.BIO.INDIANA.EDU" This archive does not support e-mail distribution directly. To use it, you must have FTP on your host For FTP, use IuBio is on the Internet network of computers with the name IuBio.Bio.Indiana.Edu (IP name--the FTP host) 129.79.1.101 (IP address) Each FTP program has it's own peculiarities, but most follow a general syntax: ftp iubio.bio.indiana.edu -- connect to iubio user: anonymous password: guest or your real user name (prefered) ? or help -- general help for ftp cd [.subdirectory] -- change directory binary -- use full binary transfer ascii -- use text transfer get any.file -- fetch a file from the archive put my.file -- put a file to the archive (only for [archive.receive] directory) bye -- close the connection Future developments: Read/Write optical drives, like the one in the NeXT computer, are becoming practical alternatives to hard disks. I recently talked to a user that had a 1000megabyte read/write drive on his mac. He said performance was comparable to a fast hard disk, and the optical disk is removable! It is likely that the price on this equipment (currently $2000-10,000) will drop soon and CD-ROM readers will become less useful, since they have very slow access times. The new low-cost macs (Classic, LC, and si) should greatly increase the number of mac users out there. With the advent of built-in sound sampling on the new macs, expect programs using this feature for sequence entry, voice recognition, etc.