The methodology and software implementation for the analysis of dose-response microarray experiment were developed over the last five years and discussed in the book "Modeling dose-response microarray data in early drug development using R". The book structure is shown in the figure below.
Although, the main part of the book is devoted to the specific setting of microarray dose-response experiment we introduce the main concept of dose-response modeling in the first part of the book. The estimations under order restrictions and inference are discussed in Chapter 2-3 while parametric non-linear modeling of dose-response data are described in Chapter 4. The methodology discussed in these chapters is introduced in a general setting and materials for these chapters are used throughout the second part of the book.
The second part of the book starts with an introduction to dose-response microarray experiments and their specific data structure in Chapter 5, in which the case study are introduced as well. The analysis of microarray data introduce the challenge of multiple testing. A general guidance for the multiple testing problem in a microarray setting is given in Chapter 6. We discuss several methods and adjusting procedures such as Bonferoni and Holm's procedures for controlling the family wise error rate and BH-FDR and SAM for controlling the false discovery rate. (Chapter 7, 8, and 12) are devoted to order restricted inference in the dose-response microarray setting. We discuss several inference procedures for detecting genes with monotone dose-response relationships such as permutation tests, significance analysis of microarray data (SAM) for dose-response data and Bayesian approaches.
More advanced inferential topics in within the dose-response microarray setting are given in Chapter 14, 15 and 16 in which we discuss the topics of multiple contrast tests, ratio tests and FDR adjusted confidence intervals. Three chapters in the book are devoted to methods which can be used in order to interprete the results obtained from the inference step. Chapter 9 and 10 present methods for classification and clustering of dose-response curves which can be applied after an initial inference step (discussed in Chapter 7 and 8) while Chapter 11 focus on the interpretation of the genes detected using the Gene Set Analysis based on Genome Ontology library. An mentioned above, we discussed the general concept of parametric dose-response modeling in Chapter 4 in the first part of the book. In Chapter 13 we focus on parametric modeling of dose-response microarray data. Note that in contrast to other chapters in the second part of the book, the aim of the analysis presented in Chapter 13 is not to detect genes with significant monotone trend but to perform secondary analysis in which characteristic of the dose-response relationship are investigated in more details.