Microarrays for the Neurosciences: An Essential Guide

Edited by:
  • Daniel H. Geschwind &
  • Jeffrey P. Gregg
MIT Press, Massachusetts, 2002. $55.00 hardcover, pp 288 ISBN 0-262-07229-7 | ISBN: 0-262-07229-7

The explosion of interest in genomic tools, particularly microarray technology, is making profound changes in how scientific questions are addressed. This growth has led to the publication of a variety of guides to help people enter the arena, including two recent books: Microarrays for the Neurosciences and DNA Arrays: Technologies and Experimental Strategies. As its title suggests, the former book is directed to neuroscientists and ranges from preparation and analysis to application and data interpretation. The latter book is a more general overview of available technologies, with particular emphasis on commercial approaches. Both books are intended for beginning to intermediate users, although individual chapters will be useful to scientists with greater experience. Both books have hefty contributions from commercial vendors of array solutions.

The great transcriptional complexity found in even the simplest neural system demands parallel and high-throughput analytical approaches. The eleven chapters in Microarrays for the Neurosciences provide a solid overview of the basics to allow scientists to quickly begin designing experiments and acquiring data. The text includes protocols as well as appendices indicating resources and suppliers. Of particular interest to biologists are chapters on scanning and data acquisition, data management and statistical issues. Authors from Axon Instruments (a leading scanner manufacturer) provide a clear description of the issues involved in maximizing signal-to-noise ratio, a critical first step in array analysis. More importantly, they explain the limits of array data for biologists who are unfamiliar with optics and instrumentation. Array data sets require a variety of software tools and databases. Several companies have developed products to complement open-source and free tools. Shah and Shams (from software-maker Biodiscovery, LLC) address spot identification, quantitation and data visualization. This chapter also introduces clustering and higher-order array analysis. Array use and technologies have led to a demand for statistical methods, both to assess the quality of an individual array and to assign confidence to the measurements. Nadon and co-authors discuss this issue and describe their statistical method in the fourth chapter. These chapters give new users a strong foundation for understanding array methods.

The final chapters of Microarrays for the Neurosciences focus more closely on the application and results of arrays in neuroscience, but are relevant to researchers in other fields as well. In particular, preparation of probes from fixed tissues and isolation of individual cells would be of interest to cancer biologists with access to tumor banks and concerned with sample heterogeneity. The heterogeneity in all tissues can mask real expression differences in a subset of cells. Unfortunately, isolation of smaller amounts of tissue or individual cells makes detection increasingly difficult, whereas amplification methods obscure quantitative differences.

DNA Arrays: Technologies and Experimental Strategies

Edited by:
  • Elena V. Grigorenko
CRC Press, Florida, 2001. $119.95 hardcover, pp 184 ISBN 0-849-32285-5 | ISBN: 0-849-32285-5

DNA Arrays: Technologies and Experimental Strategies provides a similar overview of current array technologies and approaches in the early chapters but is less successful at separating the commercial pitch from the technology in later chapters. The emphasis on individual commercial products and their application diminishes the book's general utility. Moreover, chapter 9 of this book (published first) is nearly identical to chapter 5 of the MIT Press book, including identical figures. The recycling of such material is both surprising and disappointing. Freeman and Vrana present a particularly good and comprehensive overview of the major issues, essential reading for any new user. Post-array confirmation, sample heterogeneity, sensitivity and other technical issues are laid out clearly and concisely. The book would have been greatly improved had the other authors followed this outline and addressed each issue in turn.

Tracking and analyzing data are clearly critical to microarray work. Commercial software packages are presented to address a variety of data mining and management functions, but the heavy emphasis on the authors' own products compromises the value of the effort. Grinstein and co-authors make an honest and accurate assessment of data mining: “Taken to its extreme, sick data mining looks like healthy data mining if one does not seek the advice of a professional. Shrink-wrapped software is not a solution.” This is perhaps the most important requirement to remember, highlighting the need for serious biologists to collaborate with bio-informaticians and physicists.

Any researcher who initiates arrays from scratch must overcome many technical and practical difficulties. With technological growth and development of increasingly good commercial products, the barriers to entry are becoming less technical and more financial. Many institutions have responded by instituting core facilities to provide services. For researchers interested in applying such genomic approaches to their efforts, both books offer valuable insights. Unfortunately, neither would be sufficient on its own.