Abstract

Computational molecular biology: an introductionPeter Clote and Rolf Backofen (Simon Levin, ed.), 2000, Wiley Series in Mathematical and Computational Biology, Wiley US$120 (hbk) ISBN 0 471 87251 2/US$55 (pbk) ISBN 0 471 87252 0This introductory book about computational molecular biology is aimed at advanced undergraduate and graduate students of bioinformatics, computer science, statistics, mathematics and the biological sciences. It not only introduces the reader to the commonly used algorithms in the field, but also presents the mathematics behind why they actually work.The authors use the terms ‘computational biology’ and ‘bioinformatics’ interchangeably however, most practitioners would agree that computational biology is associated with the design of algorithms, whereas bioinformatics involves the application of the algorithms to real biological problems. Going by these loose definitions, this book is clearly not a self-contained introduction to bioinformatics because the informatics problems that are encountered by researchers using very large datasets are not the focus of the book. Furthermore, there are only seven databases listed in the appendix but the number of useful online bioinformatics resources is at least triple that number!Practicalities aside, the particular strength of this book is its in-depth coverage of the mathematics that underlie the algorithms used for solving biological problems computationally. After a short chapter about molecular biology, there is a very useful math primer covering basic probability theory, information theory and combinatorial optimization. The book is then divided into chapters about sequence alignment, molecular evolution, Hidden Markov models and structure prediction.The extensive coverage of sequence alignment is invaluable to anyone wanting to understand the mathematics behind scoring matrices [including background on the well known point accepted mutation (PAM) matrices], affine gap penalties, distance measures, Smith-Waterman, Needleman-Wunsch and the Gibbs sampler. Biologically interesting applications include genomic rearrangements, locating cryptogenes and finding the expected lengths of guide RNA in Trypanosomes.The chapter about evolution begins with a motivating biological introduction and later presents evolution as a continuous Markov process. There is precise coverage of important clustering methods to generate phylogenetic trees, and the relatively rare topics of quartet trees and quartet puzzling are also included.Hidden Markov models are ubiquitous in computational molecular biology and this book devotes an entire chapter to them. The sections on evaluating models using Baum-Welch and Viterbi scores, and using expectation maximization (EM) and gradient descent to optimize transition probabilities are particularly useful. Biological applications include finding protein motifs and eukaryotic DNA promoter regions.The largest and most comprehensive chapter is dedicated to structure prediction. The introductory historical account gives a well-rounded perspective and primes the reader's interest for finding out how problems were first posed and then solved, using molecular dynamics, secondary structure prediction, homology and pattern recognition, Monte Carlo techniques, simulated annealing, constraint programming, amino acid pair potentials and protein threading.There are some important computational methods that aren't covered in detail in this book, notably genetic algorithms and neural networks. The coverage of rule-based methods is sparse, although an example of their success in finding anchor regions in RNA is cited.Each chapter of the book outlines clear and concise pseudo-code for many important algorithms, and C/C++/Java source code is available on the Wiley website. The class-tested exercises at the end of each chapter are nontrivial and thought provoking.In summary, this book is a highly technical, theoretically oriented introduction to computational molecular biology. It is best suited for its intended audience of serious readers seeking in depth coverage of the mathematics behind the computation in modern molecular biology.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call