Customer Reviews

Phylogenetics (Oxford Lecture Series in Mathematics and Its Applications)

5 star:		(0)
4 star:		(2)
3 star:		(0)
2 star:		(0)
1 star:		(0)

 

 

Starting with genes, proteins, or other biological traits, phylogenetics is about describing relationships between them. Phylogenetics tries to estimate "family trees" given only the family members visible today - exact lineage is guesswork, since the parents, grandparents, and shared heritage can never be known.

This book offers deep analysis of one family of techniques for deducing possible trees. It gives a very thorough, formal description of ways to examine and resolve different sources of information, or to determine that they can not be resolved. It offers minute analysis of ways to take subsets of the whole family, analyze the subsets, then merge the subset conclusions together, as much as possible. It also addresses the statistical character of the tree-building problem. The reader who masters this material has a powerful set of tools for phylogenetic analysis.

That reader must be truly dedicated, though. The first two chapters read like mathematical graph theory (because they are). The next few chapters are also highly mathematical, but offer a bit more biological insight. I'm not a mathematician, so I find this book tough going. The graph-theoretic conclusions give wonderful insight into combining information from multiple traits and in noting points of conflict. It takes me a while, though, to unwind the formal notation enough to attach biological meaning to it. There are a few helpful statistical analyses, but they could be missed - the more familiar kinds of statistics are hidden among the combinatorics and tree perturbations. Later chapters revisit familiar topics like parsimony and Markov models, but with theoretical depth that's hard to find elsewhere.

Within the whole gamut of phylogenetic techniques now used, this book addresses only one range. Within that range, however, Semple and Steel have done a fine job of showing the theory behind those techniques. I value the insights that this book brings. Even so, it's not always easy to dislodge those insights from the solid slabs of proofs in which they are embedded. I appreciate the demonstration of NP-completeness of specific problems, but I can't always apply that knowledge to the biology I want to address.

Anyone devoted to mastering every nuance of phylogenetic analysis should read this book. It goes beyond the needs of most application developers, though. It probably won't say much at all to those who just use the results of analysis; it simply does not address any particular application that an analyst might use. If you have the determination to understand and the patience to pick out the understanding, you'll find a lot to like in "Phylogenetics".

I agree with a lot of what the review by wiredweird said in "Good, for the dedicated reader". I just wanted to add that I believe that this book would be totally unreadable to all biologists that do not also have a fairly strong math or computer science background. As a computer scientist though, I found that this book was very useful because the explanations presented went beyond the frequently used buzzwords and included math/graph explanations of the techniques. While there is no code to copy from the book, I found most explanations sufficient enough that I could implement many of the techniques on my own if I wanted to.