Summary and Info
McLachlan is a very well-known statistician who specializes in classification, pattern recognition and mixture distribution models. I was surprised to see him write a book on microarray data. But I shouldn't have been. It turns out that in addition to data processing and statistical design, cluster analysis and classification are important aspects of the identification of genes that are really expressing themselves in an array.The book is designed for researchers who need to know a little about statistics and its role in analysis of microarray data and for statisticians who may know little or nothing about genes and microarrays. The purpose of Chapter 1 is to acquaint the statistician with the historical development of microarrays and to provide a brief tutorial to make the rest of the book more easily understood.Chapter 2 explains why microarray data needs preprocessing (cleaning and normalization) For the researcher with little familiarity with statistics important concepts and techniques are discussed in detail. The key examples are multiplicity, principal component analysis, clustering, discriminant analysis, mixture distributions, determining number of mixtures, cross-validation, classification trees, bootstrap, and selection bias.As with other books that Mclachlan authors or coauthors the book is very well-organized and well-written. It is a great resource for me and I am sure many other statisticians like me who work in medical research.