I.  What these programs do:

All widely-used methods of alignment and phylogeny inference make certain
assumptions about sequence evolution.  These assumptions may be implicit (e.g.
the parsimony method for phylogeny reconstruction, the conventional dynamic
programming algorithm for alignment inference) or they may be explicit.  The 
methods implemented here are derived from evolutionary models with explicit 
assumptions.  

This set of programs estimates (via an expectation-maximization algorithm)
parameters relevant to the evolutionary relationship between a pair of DNA 
or protein sequences.  It performs pairwise alignment and 
allows reliability of alignment positions to be examined visually.  Also, the 
phylogeny reconstruction method described in Thorne and Kishino (1992, MBE 
9:1148-1162) is implemented. 

To understand these programs a bit better and find a list of relevant 
references, read the file ``statprogs.doc''.   It may also be worthwhile
to read the files ``arcs2ps.doc'' and ``trees.doc''.  The latter two
files respectively explain the alignment visualization and phylogeny
reconstruction software.  Use these programs at your own risk !! They can 
require massive amounts of computer time (don't attempt to use them if you 
don't have a workstation or better) and they are user-rude.  A little 
familiarity with Unix, a computer that can compile C code, and ability to 
edit files are essential.  Despite these drawbacks, the methods implemented 
here can yield comparatively accurate estimates of evolutionary parameters 
and relatively high quality alignments.  

For the most part, these programs were written by Jeff Thorne with 
intellectual help from Hirohisa Kishino, Gary Churchill, and Joseph 
Felsenstein.  Sean Lamont and Kerry Neef supplied helpful programming tips.  
Leigh Thorne provided constructive comments.  Programs not written by Jeff 
Thorne are part of the phylogeny inference package (PHYLIP) distributed by 
Joseph Felsenstein.  Those programs from PHYLIP are noted in the following 
section.  
-------------------------------------------------

II. Files that you will need  ...

A.   ...if you want to read this file

->  readme (this file)

B.  ...if you want to estimate evolutionary parameters that relate a pair
of sequences or want a maximum likelihood alignment between the sequences

-> statprogs.doc (documentation for the relevant software)
-> arrayutil.c
-> commontg3.c
-> dens4.c
-> docons4.c (only when you want statprotC or statdnaC -- see statprogs.doc)
-> domask.c
-> domin4.c  (only when you want statprot or statdna -- see statprogs.doc)
-> emhtg3.c
-> gsearchtg3.c
-> inout.c
-> inv3.c
-> main.c
-> recreate3.c
-> topstg3.c

-> common.h
-> dna.h    (only for DNA sequence analysis)
-> emh4ps.h
-> prottg.h (only for protein sequence analysis)

-> evalfile (only for protein sequence analysis)
-> evecfile (only for protein sequence analysis)
-> ievecfile (only for protein sequence analysis)
-> reveqps (only for protein sequence analysis)
-> thisfar

-> dnamask.dat (only for DNA sequence analysis)
-> protmask.dat (only for protein sequence analysis)

-> makefile 

C.  ...if you also want to produce postscript output that will help visualize
    alignment reliability, you will need the files of Section B and ...

-> arcs2ps.doc (documentation for alignment-visualization software)
-> arcs2ps.c 
-> control.dat (to produce black-and-white postscript output)
-> controlcol.dat (to produce color postscript output)

D. ...if you want to infer phylogenies, you will need the files of Section B
   and ...

-> consense.c (a consensus program from PHYLIP)
-> consense.doc (documentation for the consense.c, also part of PHYLIP)
-> fitch.doc (a documentation file from PHYLIP)
-> inseed 
-> medians.c
-> modfitch35.c (a slight modification of the fitch.c program in PHYLIP)
-> phylip.h (another file from PHYLIP)
-> seqshell
-> shellprog.c
-> simdists.c
-> trees.doc (documentation for ``alignment-free'' phylogeny inference)
----------------------------------------------------------------------
III. My address

Jeffrey Thorne
Program in Statistical Genetics, Statistics Department
North Carolina State University
Raleigh NC 27695-8203

email: thorne@stat.ncsu.edu
phone: (919) 515-1946
FAX: (919) 515-7591