Machine Learning Approaches To Bioinformatics

by Zheng Rong Yang

Overview

This book covers a wide range of subjects in applying machine learning approaches for bioinformatics projects. The book succeeds on two key unique features. First, it introduces the most widely used machine learning approaches in bioinformatics and discusses, with evaluations from real case studies, how they are used in individual bioinformatics projects. Second, it introduces state-of-the-art bioinformatics research methods. The theoretical parts and the practical parts are well integrated for readers to follow the existing procedures in individual research.Unlike most of the bioinformatics books on the market, the content coverage is not limited to just one subject. A broad spectrum of relevant topics in bioinformatics including systematic data mining and computational systems biology researches are brought together in this book, thereby offering an efficient and convenient platform for teaching purposes.An essential reference for both final year undergraduates and graduate students in universities, as well as a comprehensive handbook for new researchers, this book will also serve as a practical guide for software development in relevant bioinformatics projects.

Product Details

ISBN-13:	9789814287302
Publisher:	World Scientific Publishing Company, Incorporated
Publication date:	05/07/2010
Series:	Science, Engineering, And Biology Informatics , #4
Pages:	336
Product dimensions:	6.10(w) x 9.00(h) x 0.90(d)

Table of Contents

Preface v

1 Introduction 1

1.1 Brief history of bioinformatics 3

1.2 Database application in bioinformatics 6

1.3 Web tools and services for sequence homology Alignment 8

1.3.1 Web tools and services for protein functional site identification 9

1.3.2 Web tools and services for other biological data 10

1.4 Pattern analysis 10

1.5 The contribution of information technology 11

1.6 Chapters 12

2 Introduction to Unsupervised Learning 15

3 Probability Density Estimation Approaches 24

3.1 Histogram approach 24

3.2 Parametric approach 25

3.3 Non-parametric approach 28

3.3.1 K-nearest neighbour approach 28

3.3.2 Kernel approach 29

Summary 36

4 Dimension Reduction 38

4.1 General 38

4.2 Principal component analysis 39

4.3 An application of PCA 42

4.4 Multi-dimensional scaling 46

4.5 Application of the Sammon algorithm to gene data 48

Summary 50

5 Cluster Analysis 52

5.1 Hierarchical clustering 52

5.2 K-means 55

5.3 Fuzzy C-means 58

5.4 Gaussian mixture models 60

5.5 Application of clustering algorithms to the Burkholderia pseudomallei gene expression data 64

Summary 67

6 Self-organising Map 69

6.1 Vector quantization 69

6.2 SOM structure 73

6.3 SOM learning algorithm 75

6.4 Using SOM for classification 79

6.5 Bioinformatics applications of VQ and SOM 81

6.5.1 Sequence analysis 81

6.5.2 Gene expression data analysis 83

6.5.3 Metabolite data analysis 86

6.6 A case study of gene expression data analysis 86

6.7 A case study of sequence data analysis 88

Summary 90

7 Introduction to Supervised Learning 92

7.1 General concepts 92

7.2 General Definition 94

7.3 Model evaluation 96

7.4 Data organisation 101

7.5 Bayes rule for classification 103

Summary 103

8 Linear/Quadratic Discriminant Analysis and K-nearest Neighbour 104

8.1 Linear discriminant analysis 104

8.2 Generalised discriminant analysis 109

8.3 K-nearest neighbour 111

8.4 KNN for gene data analysis 118

Summary 118

9 Classification and Regression Trees, Random Forest Algorithm 120

9.1 Introduction 120

9.2 Basic principle for constructing a classification tree 121

9.3 Classification and regression tree 125

9.4 CART for compound pathway involvement prediction 126

9.5 The random forest algorithm 128

9.6 RF for analyzing Burkholderia pseudomallei gene expression profiles 129

Summary 132

10 Multi-layer Perceptron 133

10.1 Introduction 133

10.2 Learning theory 137

10.2.1 Parameterization of a neural network 137

10.2.2 Learning rules 137

10.3 Learning algorithms 145

10.3.1 Regression 145

10.3.2 Classification 146

10.3.3 Procedure 147

10.4 Applications to bioinformatics 148

10.4.1 Bio-chemical data analysis 148

10.4.2 Gene expression data analysis 149

10.4.3 Protein structure data analysis 149

10.4.4 Bio-marker identification 150

10.5 A case study on Burkholderia pseudomallei gene expression data 150

Summary 153

11 Basis Function Approach and Vector Machines 154

11.1 Introduction 154

11.2 Radial-basis function neural network (RBFNN) 156

11.3 Bio-basis function neural network 162

11.4 Support vector machine 168

11.5 Relevance vector machine 173

Summary 176

12 Hidden Markov Model 177

12.1 Markov model 177

12.2 Hidden Markov model 179

12.2.1 General definition 179

12.2.2 Handling HMM 183

12.2.3 Evaluation 184

12.2.4 Decoding 188

12.2.5 Learning 189

12.3 HMM for sequence classification 191

Summary 194

13 Feature Selection 195

13.1 Built-in strategy 195

13.1.1 Lasso regression 196

13.1.2 Ridge regression 199

13.1.3 Partial least square regression (PLS) algorithm 200

13.2 Exhaustive strategy 204

13.3 Heuristic strategy - orthogonal least square approach 204

13.4 Criteria for feature selection 208

13.4.1 Correlation measure 209

13.4.2 Fisher ratio measure 210

13.4.3 Mutual information approach 210

Summary 212

14 Feature Extraction (Biological Data Coding) 213

14.1 Molecular sequences 214

14.2 Chemical compounds 215

14.3 General definition 216

14.4 Sequence analysis 216

14.4.1 Peptide feature extraction 216

14.4.2 Whole sequence feature extraction 222

Summary 224

15 Sequence/Structural Bioinformatics Foundation - Peptide Classification 225

15.1 Nitration site prediction 225

15.2 Plant promoter region prediction 230

Summary 237

16 Gene Network - Causal Network and Bayesian Networks 238

16.1 Gene regulatory network 238

16.2 Causal networks, networks, graphs 241

16.3 A brief review of the probability 242

16.4 Discrete Bayesian network 245

16.5 Inference with discrete Bayesian network 246

16.6 Learning discrete Bayesian network 247

16.7 Bayesian networks for gene regulatory networks 247

16.8 Bayesian networks for discovering Peptide patterns 248

16.9 Bayesian networks for analysing Burkholderia pseudomallei gene data 249

Summary 252

17 S-Systems 253

17.1 Michealis-Menten change law 253

17.2 S-System 256

17.3 Simplification of an S-system 259

17.4 Approaches for structure identification and parameter estimation 260

17.4.1 Neural network approach 260

17.4.2 Simulated annealing approach 261

17.4.3 Evolutionary computation approach 262

17.5 Steady-state analysis of an S-system 262

17.6 Sensitivity of an S-system 267

Summary 268

18 Future Directions 269

18.1 Multi-source data 270

18.2 Gene regulatory network construction 272

18.3 Building models using incomplete data 274

18.4 Biomarker detection from gene expression data 275

Summary 278

References 279

Index 319