DNA Array Image Analysis: Nuts & Bolts

Rating:
Summary: CURRENT DRUG DISCOVERY , DECEMBER 2002 - Book Review
Review: DNA Array Image Analysis is the most comprehensive book that I have read, dealing with Microarray analysis topic. My background is in genomics, which is too far away from statistics. Nevertheless, this book gave me a very nice perspective on statistical approaches for microarrays. Recommend it to all, especially for those who are running microarray facilities.

Rating:
Summary: Nuts, Bults and More for your microarrays
Review: DNA Array Image Analysis is the most comprehensive book that I have read, dealing with Microarray analysis topic. My background is in genomics, which is too far away from statistics. Nevertheless, this book gave me a very nice perspective on statistical approaches for microarrays. Recommend it to all, especially for those who are running microarray facilities.

Rating:
Summary: CURRENT DRUG DISCOVERY, DECEMBER 2002 - Book Review
Review: DNA microarrays form an indispensable tool in molecular biology, allowing researchers to gain insight into the intricate mechanisms of gene expression. Microarray technology also aids the researcher in transforming and supplementing data available on genes and cells into useful information about gene expression, and ultimately, cellular biology.
cDNA arrays experiments use many gene-specific polynucleotides derived from the ends of RNA transcripts. These are arrayed on a single matrix and simultaneously probed with a fluorescently tagged cDNA representation of total RNA pools from test and reference cells. This allows one to determine the relative amount of transcript present in the pool by the type of fluorescent signal generated. Thus, the relative message abundance is based on a comparison of the test cell state to a reference cell state. Common to all array-based technologies is the need to analyze digital images of the scanned DNA array.
Shishir Shah is an Associate Professor at Wayne State University in Detroit, and is the author of numerous publications on image analysis and data mining, while Gerda Kamberova is an Assistant Professor in Computer Science at Hofstra University, New York, where she teaches computer vision, computer graphics, and artificial intelligence. With this book, the authors address the fact that the topic of microarray image processing is usually neglected due to the availability of various software tools for image processing. The focus of microarray bioinformatics today is data analysis, but how can one be certain that the data obtained through image analysis of a microarray experiment through is of high quality? How can one perform data mining on data derived from different microarray technologies, where microarray images were analyzed with different image analysis softwares?
In Kamberova's introductory chapter, she describes the basics of image analysis for molecular biology researchers. Although some of the formulas in her chapter may seem intimidating to many biologists, they are necessary to fully understand the subject. However, the mathematical load in this book is concentrated in this chapter, and the text thereafter becomes very friendly, with more advanced discussions saved for an appendix.
Chapters on the design and performance of CCD and laser microarray scanners serve as a forum for three companies representing a good cross-section of the technology in the microarray arena. The book then moves on to spot finding and segmentation issues with tips for practical image analysis. Quality control of image analysis is discussed at the application level.
Elsewhere, Kamberova and Shah have invited a team from Fox Chase Cancer Center to discuss aspects of image background corrections and data normalization. Another interesting aspect of this publication is the inclusion of statistical comparison of data generated by various softwares. Finally, Shah's chapter on BAC arrays and image analysis is of particular interest because it is a new application gaining momentum in the microarray field.
This book emphasizes aspects of both theory and application and therefore, can be used for teaching as well as self-study. It will be a useful reference not only for computer and biology scientists, but also for anyone using or interested in microarray technology.

Rating:
Summary: Good book, great value.
Review: To my knowledge this is the first book to appear focusing on analysis of microarray images, though the topic is commonly addressed to some extent in general books on microarray technology. Because microarray technology evolves so quickly, books such as this one are almost of necessity put together in a hurry. Unfortunately, the hurry really shows in this case. In addition to the grammatical problems all too common in scientific texts, there are technical inconsistencies. For example, the term "signal to noise ratio" is defined at three places in the book, with three different definitions. First, we are told (chapter two, page 38) that the signal to noise ratio is defined as the ratio of mean signal to mean error. Later (chapter five, page 86), the signal to noise ratio is defined as signal (minus background), divided by the standard deviation of background. Finally (chapter six, page 101) tells us that the signal to noise ratio "can be estimated by computing the peak signal divided by the variation in the signal." Only that part of the discussion starting on page 86 is included in the index.

Chapter one provides an introduction to microarrays. Chapter two introduces image analysis with a focus on issues pertinent to micro array analysis, though it is not perfectly customized as background for the other chapters. For example, though Chapter two's appendixes include Fourier analysis, this is not really used explicitly further in the book. The next three chapters concern microarray scanning. They are each written by groups at different commercial firms, and they provide vendor oriented views. These chapters are redundant in the sense that they all cover the same topic, but taken as a whole, they provide some balance. It might be preferable to have an academic team write a single, more objective chapter, but it's understandable to take this approach in the interest of timeliness. Chapters six and seven are the ones that directly address image processing. Though Chapter seven is entitled "microarray data normalization", it really provides the most detailed information about image processing and analysis. Two short chapters cover a comparison of commercially available software and a brief overview of information workflow. The final chapter on bacterial artificial chromosomes is a general (and worthwhile) overview of that application, but it is not particularly focused on image analysis.

The book is concise (it took me about eight hours to read it cover to cover) and the price is right (under $30 at amazon.com). The editors and author are to be congratulated on producing a timely book, and I'm delighted that the publisher has made it so affordable. Keep up the good work.

Rating:
Summary: Good book, great value.
Review: To my knowledge this is the first book to appear focusing on analysis of microarray images, though the topic is commonly addressed to some extent in general books on microarray technology. Because microarray technology evolves so quickly, books such as this one are almost of necessity put together in a hurry. Unfortunately, the hurry really shows in this case. In addition to the grammatical problems all too common in scientific texts, there are technical inconsistencies. For example, the term "signal to noise ratio" is defined at three places in the book, with three different definitions. First, we are told (chapter two, page 38) that the signal to noise ratio is defined as the ratio of mean signal to mean error. Later (chapter five, page 86), the signal to noise ratio is defined as signal (minus background), divided by the standard deviation of background. Finally (chapter six, page 101) tells us that the signal to noise ratio "can be estimated by computing the peak signal divided by the variation in the signal." Only that part of the discussion starting on page 86 is included in the index.

Chapter one provides an introduction to microarrays. Chapter two introduces image analysis with a focus on issues pertinent to micro array analysis, though it is not perfectly customized as background for the other chapters. For example, though Chapter two's appendixes include Fourier analysis, this is not really used explicitly further in the book. The next three chapters concern microarray scanning. They are each written by groups at different commercial firms, and they provide vendor oriented views. These chapters are redundant in the sense that they all cover the same topic, but taken as a whole, they provide some balance. It might be preferable to have an academic team write a single, more objective chapter, but it's understandable to take this approach in the interest of timeliness. Chapters six and seven are the ones that directly address image processing. Though Chapter seven is entitled "microarray data normalization", it really provides the most detailed information about image processing and analysis. Two short chapters cover a comparison of commercially available software and a brief overview of information workflow. The final chapter on bacterial artificial chromosomes is a general (and worthwhile) overview of that application, but it is not particularly focused on image analysis.

The book is concise (it took me about eight hours to read it cover to cover) and the price is right (under $30 at amazon.com). The editors and author are to be congratulated on producing a timely book, and I'm delighted that the publisher has made it so affordable. Keep up the good work.