Microarray Bioinformatics

Rating:
Summary: A Good Book for Microarray Bioinformatics
Review: I rate this book a 5 star because I believe this book is one of best bioinformatics books which make it possible for the biologists to understand the bioinformatic tools inside of microarray technology. For me the most useful chapters include Sequence Databases for Microarrays, Computer Design of Oligonucleotide Probes, Normalisation, Measuring and Quantifying Microarray Variability, Analysis of Differentially Expressed Genes. As a three-years microarray user, I still get a lot information after I read this book. However, no any bioinformatic books are perfect and complete. There are also some limitations in this book. The author sometimes did not provide detailed information on some biostatistic analysis tools and only provided some references for reading. Since a lot of bioinformatic tools are still in the trial stage and need to be improved, we can not blame the author for the incompleteness.
As a 250-pages bioinformatics book, I believe, this book is very informative and useful for microarray users and biologists who are tired of understanding the abstract biostatistic equations.

Rating:
Summary: Microarrays Lite
Review: It just doesn't have the detail I wanted.

There's a lot to like here. Stekel covers everything, starting with selecting the probes and printing the arrays. Next comes raw array analysis - scanning, image processing, and measuring the effects of the array itself on the results. That covers the first six chapters. The next three go over analysis of the result, one more chapter covers experimential design, and the last chapter discusses storing, labelling, and sharing the data. Some of those topics, like experiment design, address issues that most other authors neglect.

Still, I came away feeling that I had read only half of each chapter. Going back, it turned out that I hadn't missed anything that really was there. I missed a lot, though. For example, probe selection includes a discussion of self-hybridization - good stuff. It stopped short of giving me any clear idea how much self-complementarity is too much. It mentioned DNA melting points, but without enough information for me to understand what is really melting, or how or why to choose one melting point over another. Handling of raw array data discussed Loess regression as a way to cancel out process differences across a single array. Again, it's good stuff, but what exactly is a Loess regression? Expression analysis mentions Spearman correlation as an alternative to Pearson correlation - it give Pearson's formulas, but not Spearman's. Later, when the author does give a "formula" for selecting sample sizes, it turns out to be some macro reference for some stat package. Throughout the book, I felt the same lack: I learned the names of many things, but not what they really are.

Maybe this book is OK for a first introduction. If you've had that introduction and want to take the second steps, this book probably won't meet your needs.

Rating:
Summary: an intro. for biologists
Review: This book is written clearly, which also means it doesn't touch too deep. I believe it's mainly useful for biologists who want to get a brief and application oriented introduction, but not for the researchers that want to improve the technology.