A thesis submitted to the Nanyang Technological University in partial fulfilment of the requirement for the degree of Doctor of Philosophy 2020 LIST OF FIGURES 13 LIST OF TABLES 15 1.6 Clinical Trial Results of Tamoxifen, CYP2D6 genotype, and Breast Cancer 1.7 Genome-wide scans identify true positives in complex traits 1.7.1 Type 2 Diabetes 1.7.2 Amyotrophic Lateral Sclerosis (ALS), a.k.a Lou Gehrig's disease 1.7.3 Warfarin Case Study 1.8 Relationship between CYP2D6 genotype and serum endoxifen concentrations 1.9 Implementation of genotyping in clinical testing 1.10 Other genetic association tests with tamoxifen and metabolite concentrations and metabolite ratios 1.11 Current knowledge gaps 1.12 Aims and objectives of this thesis 2 METHODOLOGY 49 2.1 Patients and Sample Collection 2.2 Measurements of steady state serum levels of tamoxifen and the active metabolite Zendoxifen 2.3 Genotyping Platforms 2.4 Software Selection for Processing Genotyping Data and Linear Regression Analysis 2.5 Quality Control of Genotyping Data 2.5.1 Determining level of missingness in the data 2.5.2 Heterozygosity 2.5.3 Population Stratification 2.5.4 SNP-level QC 2.6 Association Testing 2.6.1 Linear Regression 2.6.2 Visualising the phenotype distribution of our study cohorts 2.6.3 Fixed and Random Effects in Meta-Analysis 2.6.4 Visualisation of Association Testing 2.6.5 Genomic Inflation 2.7 Haplotype Phasing 2.8 Genotype Imputation 2.8.1 Imputation Quality Control 2.9 Processing of ChIP-seq raw reads from NCBI Sequence Read Archive (SRA) 2.10 Statistical Power Calculation 2.11 Interpretation of effect size ( ) in the log-transformed linear regression model 2.12 Additional Notes 3 RESULTS 67 3.1 Patient demographic and genetic ancestry 3.2 Data transformation for serum tamoxifen metabolite concentrations 3.3 Association testing of CYP2D6 Genetic Variants with tamoxifen metabolites 3.4 Effectiveness of CYP2D6 variants at predicting endoxifen levels 3.5 Effectiveness of Activity Score 10 3.6 GWAS of tamoxifen metabolites 3.6.1 GWAS of endoxifen 3.6.2 GWAS meta-analyses of endoxifen in breast cancer patient cohorts 3.6.3 Meta-analysis of GWAS on other tamoxifen metabolites 3.7 Fine mapping reveals index SNP rs5751245 3.7.1 Effectiveness of rs5751245 at explaining variance in comparison to CYP2D6 alleles 98 3.7.2 Conditional analysis on index SNP rs5751245 3.8 Endoxifen and N-desmethyltamoxifen SNPs 3.9 Conditional analysis on CYP2D6*4, CYP2D6*10, and CYP2D6*41 diplotype 3.10 4-hydroxytamoxifen (4OHT) Association Testing 3.11 GWAS for eQTL of select genes on Chromosome 22 3.11.1 Liver Study (SCRB) Cohort 3.11.2 eQTL GWAS results 3.12 CYP2D6 eQTL effect on endoxifen 3.12.1 eQTL analysis conditional on rs5751245 allele dosage 3.12.2 Conditional analysis on all possible CYP2D6 diplotypes 3.13 Endoxifen association testing conditional on rs2413670 allele dosage 3.14 SNP location and Open Chromatin Marks in Liver Tissue 3.15 Linkage Disequilibrium between endoxifen, CYP2D6 eQTL, and CYP2D6 allele proxy SNPs 131 3.16 Improved predictive power u...