Batch Effects and Noise in Microarray Experiments: Sources and Solutions (Wiley Series in Probability and Statistics)

Batch Effects and Noise in Microarray Experiments......Page 4
Contents......Page 8
List of Contributors......Page 16
Foreword......Page 20
Preface......Page 22
1 Variation, Variability, Batches and Bias in Microarray Experiments: An Introduction......Page 24
2.1 Introduction......Page 28
2.2.1 Affymetrix......Page 29
2.2.3 Illumina......Page 30
2.2.5 Spotted Microarrays......Page 31
2.3.2 Sample and RNA Extraction......Page 32
2.3.3 Amplification......Page 35
2.3.5 Hybridization......Page 36
2.3.6 Washing......Page 37
2.3.7 Scanning......Page 38
2.3.8 Image Analysis and Data Extraction......Page 39
2.4 Conclusions......Page 40
3.1 Introduction......Page 42
3.2.1 Definitions......Page 43
3.2.3 Biological Variation......Page 44
3.2.6 Experimental Factors......Page 45
3.2.7 Statistical Errors......Page 46
3.3 Measures to Increase Precision and Accuracy......Page 47
3.3.3 Replication......Page 48
3.3.4 Further Measures to Optimize Study Design......Page 49
3.4.2 Observational Bias......Page 51
3.4.3 Bias at Specimen/Tissue Collection......Page 52
3.5 Conclusion......Page 53
4.1 Introduction......Page 56
4.2 A Statistical Linear Mixed Effects Model for Microarray Experiments......Page 58
4.2.2 Examples of Design Guided by the Linear Model......Page 60
4.3.2 Incomplete Block Designs......Page 62
4.3.3 Multiple Batch Effects......Page 63
4.4 Reducing Batch Effects by Normalization and Statistical Adjustment......Page 64
4.4.1 Between and Within Batch Normalization with Multi-array Methods......Page 66
4.4.2 Statistical Adjustment......Page 69
4.5.1 Sample Pooling......Page 70
4.5.2 Sample Splitting: Technical Replicates......Page 71
4.7 Conclusions......Page 72
Acknowledgements......Page 73
5.1 Introduction......Page 74
5.2.1 Same Protocol, Different Times of Processing......Page 75
5.2.2 Same Protocol, Different Sites (Study 1)......Page 76
5.2.3 Same Protocol, Different Sites (Study 2)......Page 78
5.2.4 Batch Effect Characteristics at the Probe Level......Page 80
5.3 Conclusion......Page 83
6.1 Introduction......Page 84
6.1.1 Spike-in Experiments......Page 85
6.1.3 The IC Curve and MA Plot......Page 86
6.2 Pre-processing......Page 87
6.2.2 Background Correction......Page 88
6.2.3 Saturation......Page 90
6.2.4 Normalization......Page 91
6.2.5 Filtering......Page 93
6.3.2 Cluster Analysis......Page 94
6.4 Conclusion......Page 96
7.1 Introduction......Page 98
7.1.1 Microarray Gene Expression Data......Page 99
7.1.2 Analysis of Variance in Gene Expression Data......Page 100
7.3.1 Data Description......Page 101
7.3.2 Normalization......Page 102
7.4 Application: The MAQC Project......Page 104
Acknowledgements......Page 108
8.1 Introduction......Page 110
8.3.1 Assessment of Smooth Bias in Baseline Expression Data Sets......Page 112
8.3.2 Relationship between Smooth Bias and Signal Detection......Page 114
8.3.3 Effect of Smooth Bias Correction on Principal Components Analysis......Page 115
8.3.4 Effect of Smooth Bias Correction on Estimates of Attributable Variability......Page 117
8.3.5 Effect of Smooth Bias Correction on Detection of Genes Differentially Expressed by Fasting......Page 118
8.3.6 Effect of Smooth Bias Correction on the Detection of Strain-Selective Gene Expression......Page 119
8.4 Discussion......Page 120
Acknowledgements......Page 122
9.1 Introduction......Page 124
9.3 Probe-by-Probe Modeling of the Input Mass Effect......Page 126
9.4 Further Evidence of Batch Effects......Page 131
9.5 Conclusions......Page 133
10.1 Introduction......Page 136
10.1.1 Bayesian and Empirical Bayes Applications in Microarrays......Page 137
10.2.2 Batch Effect Adjustment Methods for Large Sample Size......Page 138
10.2.3 Model-Based Location and Scale Adjustments......Page 139
10.3.1 Parametric Shrinkage Adjustment......Page 140
10.3.2 Empirical Bayes Batch Effect Parameter Estimates using Nonparametric Empirical Priors......Page 143
10.4.1 Microarray Data with Batch Effects......Page 144
10.4.3 Results for Data Set 2......Page 147
10.4.4 Robustness of the Empirical Bayes Method......Page 149
10.4.5 Software Implementation......Page 150
10.5 Discussion......Page 151
11.1 Introduction......Page 154
11.2.1 Data Description......Page 156
11.2.2 Empirical Bayes Method for Batch Adjustment......Page 157
11.3.1 Removal of Cross-Experimental Batch Effects......Page 158
11.3.2 Removal of Within-Experimental Batch Effects......Page 159
11.3.3 Removal of Batch Effects: Empirical Bayes Method versus t-Test Filter......Page 160
11.4.1 Methods for Batch Adjustment Within and Across Experiments......Page 161
11.4.3 Implications of Cross-Experimental Batch Corrections for Clinical Studies......Page 162
12.1 Introduction......Page 164
12.2.1 Principal Components Analysis......Page 166
12.2.3 Principal Variance Components Analysis......Page 168
12.3.1 A Transcription Inhibition Study......Page 169
12.3.3 A Hepato-toxicant Toxicity Study......Page 170
12.4.1 PVCA Provides Detailed Estimates of Batch Effects......Page 171
12.4.2 Visualizing the Sources of Batch Effects......Page 172
12.4.3 Selecting the Principal Components in the Modeling......Page 173
12.5 Discussion......Page 176
13.1 Introduction......Page 178
13.2 Mouse Lung Tumorigenicity Data Set with Batch Effects......Page 180
13.2.1 Batch Profile Estimation......Page 182
13.2.2 Batch Profile Correction......Page 183
13.2.3 Batch Profile Scoring......Page 184
13.2.4 Cross-Validation Results......Page 185
13.3 Discussion......Page 187
Acknowledgements......Page 188
14.1 Introduction......Page 190
14.2 Analysis of the NCI 60 Data......Page 192
14.3 Improved Statistical Power......Page 197
14.4 Gene-by-Gene versus Multivariate Views......Page 201
14.5 Conclusion......Page 204
15.1 Introduction......Page 206
15.2 Aggregated Expression Intensities......Page 208
15.3 Covariance between Log-Expressions......Page 209
15.4 Conclusion......Page 212
Acknowledgements......Page 213
16.1 Introduction......Page 214
16.2 Potential Sources of Spurious Associations......Page 215
16.2.1 Spurious Associations Related to Study Design......Page 217
16.2.3 Spurious Associations Caused by Genotype Calling Errors......Page 218
16.3.1 Batch Effect in Genotyping Experiment......Page 219
16.3.2 Batch Effect in Genotype Calling......Page 220
Disclaimer......Page 224
17.1 Introduction......Page 226
17.3 Systems-Biological Concepts in Medicine......Page 227
17.5 Strategies for Gene Expression Biomarker Development......Page 228
17.5.1 Phase 1: Clinical Phenotype Consensus Definition......Page 229
17.5.2 Phase 2: Gene Discovery......Page 230
17.5.4 Phase 4: Diagnostic Classifier Development......Page 232
17.5.5 Phase 5: External Clinical Validation......Page 233
17.5.6 Phase 6: Clinical Implementation......Page 234
17.5.7 Phase 7: Post-Clinical Implementation Studies......Page 235
17.6 Conclusions......Page 236
18.1 Introduction......Page 238
18.2 Reporting Standards......Page 239
18.3.1 The Microarray Gene Expression Data Society......Page 242
18.3.4 The Genomic Standards Consortium......Page 243
18.3.6 Data Standards in Biopharmaceutical and Clinical Research......Page 244
18.3.7 Standards Integration Initiatives......Page 245
18.3.10 FuGE and ISA-TAB......Page 246
18.4 Experimental Standards: Developing Quality Metrics and a Consensus on Data Analysis Methods......Page 249
18.5 Conclusions and Future Perspective......Page 251
References......Page 254
Index......Page 268