Bioinformatics: Second Ed (P): Sequence and Genome Analysis [Englisch] [Taschenbuch]

The book begins with an historical introduction to the subject, and a newcomer to the subject will get a brief overview of some of the first sequence analysis programs and some of the first DNA sequence databases developed long before bioinformatics was recognized as a real discipline. The author introduces some of the techniques that will be discussed in the book, such as global and local sequence alignment, dynamic programming, RNA structure prediction, and protein structure prediction. This is followed in chapter 2 by an overview of the procedures used to collect and store sequences in the laboratory. To the reader not familiar with these techniques, the discussion may be too brief. The different sequence formats used are outlined, as well as techniques used to convert from one sequence format to another.

Chapter 3 takes a closer look at the pairwise alignment of sequences, and the author also outlines the reasons behind examining sequence alignment in the first place, namely that of finding the functional, structural, and phylogenetic information. The view of sequence alignment as an optimization problem is emphasized via the dynamic programming algorithm for sequence alignment. Dot matrix analysis is discussed a sequence alignment strategy that allows all possible matches of residues between two sequences. The author is careful to note that local alignment algorithms might give global alignments, and vice versa, because of small changes in the scoring system. The PAM and BLOSUM substitution matrices are compared as to their relative merits and pitfalls. A very detailed discussion of gap penalties is given, along with the role of the Gumbel extreme value distribution in the determination of the statistical significance of a local alignment score between two sequences. And, after a brief introduction to Bayesian statistics, the author shows how to to use it produce alignments between pairs of sequences and to calculate distances between sequences. The Bayes block aligner software package is discussed in detail as a tool for Bayesian sequence alignment.

In chapter 4, the author gives an extensive discussion of multiple sequence alignment algorithms, the most important of these by contemporary standards being hidden Markov models. The author though does treat the "progressive" methods, as well as the use of genetic algorithms in doing multiple sequence alignment. The former include the classic CLUSTALW package and the PILEUP program for doing msa. Although the discussion of hidden Markov models makes sparing use of mathematics, is does serve to explain how they work and should assist readers who need a solid understanding of them.

I did not read chapters 5 and 6 so I will omit their review. Chapter 7 is an introduction to database searches in order to find similar sequences. The algorithms developed in chapters 3 and 4 again make their appearance, and the reader is confronted with various user interfaces for performing genetic database searching online. The FASTA and BLAST tools are introduced as fast methods to do database searching. As computer performance increases in the years ahead, these and other currently existing tools will no doubt be replaced by more powerful search routines. While perusing this chapter, one cannot help but be fascinated by the current situation in the biological/genetic sciences. Once thought of as a purely descriptive science, it is now dominated by a reductionist philosophy, involving huge amounts of data, and sophisticated mathematics for the analysis of this data.

The author moves on to the methods for detecting protein-encoding regions of DNA sequences in chapter 8. The simplest method according to the author for doing this is to search for ORFs, and he discusses the reliability of methods for accomplishing this. Hidden Markov models again make their appearance as a tool to study eukaryotic internal exons and in gene prediction in microbial genomes. And, neural networks are introduced as tools for finding complex patterns and relationships among sequence positions, and Grail II is discussed as a system for exon finding in eukaryotic genes. Promotor prediction in E. Coli is also briefly overviewed.

I did not read chapter 9 so I will omit its review. Chapter 10 though is an introduction to one of most interesting parts of bioinformatics, namely that of analyzing the entire genomes of organisms. Due to rapid experimental advances in genetics, several genomes are now available, and this allows a more global, dynamical view of the role of genes and how their expression correlates to result in a fully-developed functioning organism. The techniques discussed in earlier chapters come into play in genomic analysis, and many other more novel techniques will have to be invented if sense is to be made of the enormous amount of genomic data currently available.

While this is a book by and for Biologists, I have found the book to be very rough and in need of extensive editing. On the first test the professor was disappointed as only 1-2 people made it into the A range. At the time I wondered if it was related to the difficulty of the text. To my surprise, my professor began to give us 10 to 20 page handouts per class, covering the material in his lectures. Although he never directly stated this, the handouts were apparently there to make up for the weakness of the text.

It definitely has potential to be a good text for biologists. If the author and editor put in significant work, this could really become a very good book. However I really can't recommend the current edition.