BINARY: A MAGAZINE ON COMPUTERS IN MICROBIOLOGY Editors: JWT Wimpenny (Cardiff, UK) HR Bungay (Rensselaer Polytechnic Institute New York, USA) Binary is an international journal which publishes a broad range of articles related to all aspects of computing as applied to microbiology. Each issue contains specialist refereed articles, review articles, general interest items, and reviews of software, hardware and books. A News section keeps the reader up-to-date on the latest software, hardware, books, journals, electronic communications, forthcoming meetings and events, etc.and a Literature Update lists the latest publications on offer in the field of computing in microbiology. Short programs or program extracts will be published; longer programs will be kept as listings or on disks and will available separately. Binary aims to provide a forum for information exchange between microbiologists, globally, who use computers, and is designed to appeal to a wide range of computer expertise amongst its readership, from novice to expert. Research Areas include % Biotechnology, including interfacing computers with bioreactors, laboratory automation and management and simulation/modelling of culture systems % Image analysis % Neural networks and artificial intelligence in microbiology % Computer-assisted learning/computer-based training in microbiology % Clinical microbiology, including the monitoring and control of hospital infection, detection/identification of pathogens and patient management % Microbial genetics and molecular biology, including sequence analysis % Microbial physiology, biochemistry and ecology % Microbial systematics and taxonomy, including computer-assisted data analysis, and database management for microbial strain/culture collections % Statistics and mathematical modelling % General aspects of computing, including data processing, analysis and presentation, word-processing, desktop publishing, information management, communications, and networks. SUBSCRIPTION INFORMATION: Volume 5 (6 issues) in 1993 Submissions and subscription information from the editorial office at the School of Pure & Applied Biology University of Wales College of Cardiff PO Box 915 Cardiff CF1 3TL, UK (Tel: 0222 874000 x 5743/4974; fax: 0222 874305; email: sabjbe@uk.ac.cardiff.thor BINARY - Computing in Microbiology Volume 5 Issue 4 August 1993 AU TI News. SO Binary-Comput-Micro. 1993 Aug. 5(4). P 109. AU TI Literature Update. SO Binary-Comput-Micro. 1993 Aug. 5(4). P 110. AU Dempsey-M-J. TI Software Reviews: Design of Experiments Pack Version 2. SO Binary-Comput-Micro. 1993 Aug. 5(4). P 111-113. AU Manfio-G-P. TI Software Reviews: GelManager 1.00, GelManager for Windows: gel fingerprinting analysis. SO Binary-Comput-Micro. 1993 Aug. 5(4). P 114-116. AU Muhlemann-H-M. TI Teaching Programs: Computer Tutorial about Yeast. SO Binary-Comput-Micro. 1993 Aug. 5(4). P 117-118. AU Castrillo-J-I. Ugalde-U-O. TI Mathematical Modelling of Yeast Energy Yielding Metabolism: Significance of Coupling Between Glycolytic and Oxidative Fluxes. SO Binary-Comput-Micro. 1993 Aug. 5(4). P 119-125. AB A program has been constructed for the prediction of the relative contributions of different energy yielding metabolic pathways of yeasts in steady state chemostat cultures. The program is based on a model of energy metabolism which is separated into two subunits: a) glycolytic subunit, b) respiratory subunit, through which the substrate C-flux (gram- atom g-1 h-1) is calculated. The coupling of fluxes assumes the existence of respiratory pattern. Oxidoreductive metabolism appears as a consequence of an overflow effect resulting from relative increases in the substrate flux through the glycolytic subunit with respect to the oxidative subunit, independently of the maximum respiratory capacity of the yeast. The program has been applied to the growth of Kluyveromyces marxianus in deproteinated whey. The model is able to reproduce the behaviour of K. marxianus in chemostat cultures. The program can be applied to other yeasts and conditions with minimal modifications. AU Millership-S. TI Use of a Neural Network for Analysis of Bacterial Whole Cell Protein Fingerprints. SO Binary-Comput-Micro. 1993 Aug. 5(4). P 126-131. AB Bacterial fingerprinting by electrophoresis of cellular components is a useful technique for both taxomonic and typing purposes. However, its value is limited by the difficulty of obtaining reproducible results between experiments. Neural nets are fault tolerant allowing for noisy, incomplete or erroneous data. They may offer a solution to the problem of reproducibility. A further advantage is the ability to identify a known pattern rapidly, which would be especially useful where many fingerprints must be examined. This report examines the use of a back propagation network as a means of classifying bacterial fingerprints produced by electrophoresis of whole cell proteins. Input data to the net were the amplitudes corresponding to the @I(n) lowest frequencies of the Fourier transform of a one dimensional plot of distance along an electrophoresis track against density. The net was capable of reliably discriminating molecular weight marker tracks from those of bacterial proteins, which would be of value for automated gel analysis. It could also discriminate fingerprints from different genera, and even superficially similar fingerprints within a single genus. Neural nets may be useful for rapid identification and classification, especially where many fingerprint patterns are possible or many samples must be examined. AU Barker-G-C. Grimson-M-J. TI A Cellular Automaton Model of Microbial Growth. SO Binary-Comput-Micro. 1993 Aug. 5(4). P 132-137. AB A cellular automaton model of diffusion controlled, two compartment, population growth is intoduced. The microscopic variables of the automaton may be interpreted, precisely, in terms of the macroscopic properties of a real system. The behaviour of the model system has many features in common with the growth of bacterial populations. The automaton identifies reaction-limited and diffusion-limited growth regimes and it illustrates structured cluster growth. AU Gyllenberg-M., Gyllenberg-H-G. Koski-T. Schindler-J. TI Nonuniqueness of Numerical Taxonomic Structures. SO Binary-Comput-Micro. 1993 Aug. 5(4). P 138-144. AB The most important methods of numerical taxonomy in microbiology are based on so called reference matrices giving the frequencies of positive binary features of the different taxa. Microbiologists seem to have been tacitly assuming that every well-defined classification method, that is, every algorithm for constructing a reference matrix from data, leads to a unique classification (reference matrix). We use a mathematical result Q that a finite mixture of multivariate Bernoulli distributions is always unidentifiable Q to disprove this assumption. We show that the same classification method applied to the same data can always lead to different classifications. This result is of importance for the foundations of computational microbial taxonomy. It is illustrated by simple examples from the two main methods of classification and identification: the one where classification is performed first and then followed by identification, and cumulative classification where classification and identification are carried out simultaneously. The consequences of the nonuniqueness result for microbiological practice are discussed. BINARY (1993) Volume 5 Issue 3 June 1993 CONTENTS BINARY (1993) 5(3) ISSN0266-304X AU TI News. SO Binary-Comput-Micro. 1993 Jun. 5(3). P 73. AU TI Literature Update. SO Binary-Comput-Micro. 1993 Jun. 5(3). P 74. AU Fry-J-C. TI Software Reviews: Towards an Ideal Scientific Graphic System for IBM AT Compatible Computers. SO Binary-Comput-Micro. 1993 Jun. 5(3). P 75-78. AU Wilkins-M-F. Morris-C-W. Boddy-L. TI Software Reviews: A Review of Two Professional Neural Network Development Packages Q NeuralWorks and AutoNet WorkBench. SO Binary-Comput-Micro. 1993 Jun. 5(3). P 79-81. AU Collins-M-A. TI Software Reviews: Windows Database Reviews Part I Q Approach Version 2.0 and Skylight Version 1.1. SO Binary-Comput-Micro. 1993 Jun. 5(3). P 82-85. AU Morgan-P. TI Software Reviews: Papyrus Bibliography System, Version 7.0. SO Binary-Comput-Micro. 1993 Jun. 5(3). P 86-87. AU Boddy-L. TI Book Review: Naturally Intelligent Systems. SO Binary-Comput-Micro. 1993 Jun. 5(3). P 87. AU Rouch-D. Littlejohn-T. TI The Future of Academic Biocomputing Facilities. SO Binary-Comput-Micro. 1993 Jun. 5(3). P 88-89. AU Mowry-M.C. TI Teaching Programs: Computer Tutorial on Mutation/Selection. SO Binary-Comput-Micro. 1993 Jun. 5(3). P 90. AU Hartman-D-W. TI Teaching PInspiration for Home Brewing. SO Binary-Comput-Micro. 1993 Jun. 5(3). P 91. AU Paul-G-C. Kent-C-A. Thomas-C-R. TI An Image Processing Method for the Fully Automatic Measurement of Vacuoles in Filamentous Fungi. SO Binary-Comput-Micro. 1993 Jun. 5(3). P 92-99. AB An image processing algorithm is described for a fully automatic method for the characterization of vacuolation in filamentous fungi grown in dispersed form in submerged culture. The technique allows size and shape measurements on vacuoles and empty cells, and on cells completely filled with cytoplasmic materials (healthy cells). Proportions of these three regions are quantified. The method has been tested on a laboratory scale Penicillium chrysogenum fed-batch fermentation. Analysis of a typical sample takes about 30 minutes. The basic image processing operation is detection of objects from the grey image at two greyness levels, one corresponding to vacuoles, the other to cytoplasm. Many adventitious objects (e.g. solid particles, debris, air bubbles, etc.) are often included in the detected images. A number of image processing filters using size, shape and greyness level criteria are employed to eliminate these undesired objects and artefacts. This paper describes how these filters are constructed, and how they can be successfully applied even in solids-containing medium fermentations with little or no manual intervention (using the optional manual editors provided). This image analysis technique provides a powerful tool for the quantitative characterisation of vacuolation and will allow greater understanding of physiological processes occurring during filamentous fermentations. AU Bazin-M-J.. Sanges dUAbadie-M-C. Balyuzi-M-H. TI Monte Carlo Simulation of Bacterial Growth. SO Binary-Comput-Micro. 1993 Jun. 5(3). P 100-103. AB A Monte Carlo simulation which generates a steady state frequency distribution of cell volumes based on a linear model of cell growth has been developed. Neonatal and incipient cell volumes are assumed to be normally distributed. The means and standard deviations of these distributions can be specified. The simulation is written in Pascal. Therefore, it allows the appropriate characteristics of a population generated from a single cell to be recorded. The resulting frequency distribution was compared with that derived analytically. Good agreement was obtained. AU Kalmakoff-J. Lousberg-M. TI Simulation Modelling with Hypercard. SO Binary-Comput-Micro. 1993 Jun. 5(3). P 104-108. AB Many scientists use computers to calculate or model phenomena and are comfortable using a programming language. However, to keep programming to a minimum, background information, formulae and assumptions, and a friendly user- interface are usually lacking. Yet such models can just as easily be presented in an environment where assumptions are made explicit and where exploration of the model is encouraged, without increasing the programming burden. To demonstrate this a simulation which was written in BASIC for an MS DOS based machine, previously presented in Binary 4, 103-105, 1992 was converted to Hypercard on the Macintosh. This model was expanded in various ways, resulting in an easy to use, accessible and information rich computer simulation. ---------------------------------------------------------------------- sabjw@uk.ac.cardiff.thor).