To harness the high-throughput potential of DNA microarray technology, it is crucial that the analysis stages of the process are decoupled from the requirements of operator assistance. Microarray Image Analysis: An Algorithmic Approach presents an automatic system for microarray image processing to make this decoupling a reality. The proposed system integrates and extends traditional analytical-based methods and custom-designed novel algorithms.
The book first explores a new technique that takes advantage of a multiview approach to image analysis and addresses the challenges of applying powerful traditional techniques, such as clustering, to full-scale microarray experiments. It then presents an effective feature identification approach, an innovative technique that renders highly detailed surface models, a new approach to subgrid detection, a novel technique for the background removal process, and a useful technique for removing "noise." The authors also develop an expectation–maximization (EM) algorithm for modeling gene regulatory networks from gene expression time series data. The final chapter describes the overall benefits of these techniques in the biological and computer sciences and reviews future research topics.
This book systematically brings together the fields of image processing, data analysis, and molecular biology to advance the state of the art in this important area. Although the text focuses on improving the processes involved in the analysis of microarray image data, the methods discussed can be applied to a broad range of medical and computer vision analysis areas.
Table of Contents
Current state of art
Contribution to knowledge
Structure of the book
Copasetic microarray analysis framework overview
Image transformation engine
Structure Extrapolation I
Pyramidic contextual clustering
Structure Extrapolation II
Image layout—master blocks
Feature Identification I
Evaluation of feature identification
Evaluation of copasetic microarray analysis framework
Feature Identification II
Proposed approach—subgrid detection
Chained Fourier Background Reconstruction
A new technique
Experiments and results
Graph-Cutting for Improving Microarray Gene Expression
Experiments and results
Stochastic Dynamic Modeling of Short Gene Expression Time Series Data
Stochastic dynamic model for gene expression data
An EM algorithm for parameter identification
Conclusions and future work
Contributions to microarray biology domain
Contributions to computer science domain
Future research topics
Appendix A: Microarray Variants
Appendix B: Basic Transformations
Appendix C: Clustering
Appendix D: A Glance on Mining Gene Expression Data
Appendix E: Autocorrelation and GHT
Karl Fraser is a research fellow in the Centre for Intelligent Data Analysis at Brunel University.
Zidong Wang is a professor of dynamical systems and computing in the Department of Information Systems and Computing at Brunel University.
Xiaohu Liu is a professor of computing and head of the Centre for Intelligent Data Analysis at Brunel University.
Overall, this is a well-written book, and it should be useful for researchers and practitioners who work on microarray image analysis.
—Peihua Qiu, Technometrics, May 2012