Edward Keedwell

IMHO, this book is not bad enough to be totally ignored, but it is not good enough for me to spend my money on it.

This brief book provides an overview of bioinformatics and the application of AI techniques to its problems. Personally, and as a computer scientist and an AI researcher with little background in biology, I found the chapters on the basics of molecular biology (Chapter 1), the challenges of bioinformatics (Chapter 2), genetic programming (Chapter 9), and cellular automata (Chapter 10) to be the most interesting/informative ones.

On the negative side, there are a considerable number of typos that are quite easy to spot it makes you question the editing process. Furthermore, the discussion of computational complexity classes (Section 3.9) contains some false statements. For example, it is claimed towards the end of the first paragraph on page 97 that "if a polynomial-time solution can be found for just one of the problems in the NP class, there will be polynomial-time solutions for all problems in the NP class." This could be another typo where the first "NP" should have been "NP-complete." Another example is the statement that "so far no one has managed to conclusively prove that a single problem in NP actually falls in the class P." This statement is false because the class P is contained in the class NP. In other words, there are many problems in NP that fall in P, namely all the problems of P.

Bottom line, if you are looking for a "starting point" for the application of AI techniques to bioinformatics, this book is not bad. Then again, Wikipedia is not a bad starting point either.… (mehr)