In this dissertation, the two main mechanisms of adverse event occurrence namely drug-gene and drug-drug interaction were studied to identify the cause of CER induced rhabdomyolysis. Genetic variation in drug metabolizing enzymes and membrane transporters as well as other drugs can modulate the beneficial, as well as the deleterious, effects of drugs. In a study of 126 patients who developed rhabdomyolysis while taking the HMG-CoA reductase inhibitor, cerivastatin, we sought to identify genetic variants and drug-drug interactions that might explain the high incidence of rhabdomyolysis. We re-sequenced three transporter genes, ABCC2 (coding for MRP2), ABCG2 (coding for BCRP) and SLCO1B1 (coding for OATP1B1), three metabolizing enzyme genes CYP2C8, UGT1A1 and UGT1A3 and HMGCR involved in transport, metabolism and target of cerivastatin, respectively. A total of 203 SNPs were identified in these samples and of these 52 were in the coding region.

In a previously published case-control analysis of polymorphisms identified in our CYP2C8, SLCO1B1, UGT1A1 and UGT1A3 genes, the V174A SNP was found to be significantly associated with CER induced rhabdomyolysis with an odds ratio of 1.89 (95% CI,1.40-2.56). We were able to only complete in in vitro functional analysis of variants in SLCO1B1 gene on the uptake of cerivastatin in HEK293/frt cells stabely expressing SLCO1B1 reference, polymorphisms and haplotypes. The V174A SNP, along with R57Q, P155T, FS and OATP1B1*15 and N1 haplotypes were shown in in vitro assays to be associated with significant reduction (P>0.001) in CER uptake (32%, 17.9%, 72%, 3.4%, 2.1% and 5.7% of reference, respectively) compared to reference. Furthermore, clopidogrel and rofecoxib, previously identified in our cases to be associated with cerivastatin induced rhabdomyolysis at odds ratio of 29.6 (95% CI, 6.1-143) and 4.9 (95% CI, 1.1-20.8), were shown in vitro to have a significant OATP1B1 mediated interaction, inhibiting cerivastatin uptake with IC50 values of 0.32 µM (95% CI, 0.06-1.71) and 0.73 µM (95% CI, 0.3-1.8), respectively. The calculated R value for clopidogrel and rofecoxib were greater than 2, supporting a further clinical evaluation of this drug interaction.

Drug response is well documented to vary considerably among patient groups and populations, as well as within individual patients. Since drug prescribing is often based on population averages of drug response, many patients will not respond, and up to one-third may experience harmful toxicity. Genetics plays a large role in explaining the variability observed in response to different drugs and is an important factor driving precision medicine initiatives. Pharmacogenetic information can be useful in optimizing patient therapy, potentially reducing the cost of hospitalizations and treatment of adverse drug events.

As part of the Kaiser Permanente Research Program on Genes, Environment, and Health (RPGEH), we analyzed 102,979 members of the Genetic Epidemiology Research on Adult Health and Aging (GERA) cohort with genetic information available, along with almost two decades of electronic health record (EHR) data, prescription records, and lifestyle survey results. In one of the largest, most ethnically diverse pharmacogene characterization studies to date, we assessed cohort metabolizer status phenotypes for 7 drug-gene interactions (DGIs) for which there is moderate to strong evidence suggesting the use of pharmacogenetic information to guide therapy. 89% of the cohort had at least one actionable allele for the 7 DGIs in this study, and we observed large variations among ethnicities. Additionally, 17,747 individuals had been prescribed a drug for which they had an actionable or high-risk metabolizer status phenotype. For these individuals, the availability of pharmacogenetic information at point-of-care may have potentially led to a more personalized drug or dosing regimen. Following this study, we assessed the utility of this resource for deriving two drug response phenotypes: weight gain induced by atypical antipsychotic use and major adverse cardiovascular events in clopiodgrel non-responders. Despite challenges in deriving phenotypes from the EHR, we were able to extract phenotypes that reflected observed estimates from previously published studies. Using these phenotypes, we performed candidate gene and genome-wide association studies to identify genetic variants associated with response. Altogether, this dissertation demonstrates the potential utility and clinical impact of integrating genetic data with EHRs for pharmacogenetic application and discovery, and provides the foundation for future studies in precision medicine.

Kidney transplantation recipients face rejection despite anti-rejection drugs and matching efforts. Biopsy confirmed acute rejection (AR) and chronic allograft nephropathy (CAN) are 2 rejection phenotypes of interest. We conducted a genome-wide association study (GWAS) in European-derived kidney transplant donors and recipients. Well-functioning transplant donors (TX; N=261) were compared to AR (N=90) or CAN (N=105) participants. The same comparisons were conducted in recipients (TX N=226; AR N=71; CAN N=105). Analyses were adjusted for multiple comparisons and additionally for population substructure by including the first 2 multi-dimensional scaling dimension values as covariates in logistic regression. The most significant findings were identified in the TX vs. CAN tests (lowest unadjusted P=6.51E^-08), which also displayed the largest odds ratios, and the least significant findings were identified for the TX vs. AR tests. Results need to be validated in an independent collection.

Proportion of identity by state, pi-hat, was calculated between donor-recipient pairs and compared between different donor-types and also by outcome (TX, AR, CAN). No significant difference within donor-type matched pairs was observed between outcome phenotypes in European-derived samples.

A set of primers was developed to sequence 112 candidate genes for AR and CAN. A custom resequencing tiling array was designed and tested. Technological development in the field called for testing the same panel on next-generation sequencing technology. We improved quality control metrics by trimming the reads and successfully called single nucleotide polymorphisms (SNPs) at a rate of 1/1000 bases sequenced.

Finally, TNFAIP3, a candidate gene for autoimmune disease (AID) was sequenced in samples multiply affected with AID (N=123) and in controls (N=397). One novel intronic insertion/deletion polymorphism was significantly associated with multiple AID diagnoses (Fisher's Exact P-value=0.0090; OR (95% CI) 7.053(1.67-29.79). Coding polymorphism rs2230926 was tested for association in a panel of individuals from families with multiple AIDs. Significant association was observed with all affected individuals (P=0.0336) as well as psoriasis, Crohn's disease and rheumatoid arthritis, with marginal association for Sjogren's and Graves disease. Additionally, we conducted an association study of the entire gene locus in lupus and identified 3 independent signals of association, including coding SNP rs2230926.

Understanding DNA sequence variation is the first step to understanding the underlying genetic architecture of a complex trait, often a manifestation of joint contributions from multiple genes. While instrumental to the successful completion of the Human Genome Project and still considered to be the gold standard, Sanger sequencing is rapidly being phased out for discovery as newer technologies that deliver sequence data with higher throughput and at lower costs become available. With the rise of these newer technologies, generation of sequencing data is no longer the bottleneck. However, one is faced with having to decide between competing technologies, and interpretation of sequencing data remains a significant challenge. In this dissertation, I will present my research on three sequencing projects and one genome mapping project. The first sequencing project involves the use of an array-based resequencing platform, the Affymetrix MitoChip to resequence the mitochondrial DNA for a pancreatic cancer case-control study and for a subset of samples from the Health ABC study. Array-based resequencing represents a high-throughput option for resequencing selected, well-defined regions. We uncovered many novel, rare variants in the mitochondrial sequence, some of which are highly conserved, protein-altering, and predicted to be damaging. While their functional consequences are not clear, we show that more rare variants remain to be found, and the abundance of them suggests that mitochondrial variants could contribute to disease phenotypes. Then, two targeted sequencing studies will be described. The first used an emulsion PCR based approach to study variants in genes involved in drug pathways; the second used a solution-phase target capture approach to fine map causative variants in a region in canine chromosome 6 associated with adult-onset deafness in Border collies. Finally, I will describe our use of nanochannel arrays for genome mapping, an improved version of optical mapping, in structural variation analysis and sequence assembly. Sequence motif maps were constructed 95 BACs previously sequenced for the MHC Sequencing Consortium. The BAC set was sequenced and assembled de novo with assistance from the sequence motif scaffold. We show that the sequence motif map could serve as a scaffold, facilitating gap closing and assembly finishing. Together, our work highlights the potential roles these new technologies play in genetic studies and the importance of taking a multigene approach while considering polygenic traits.

The genes encoding extracellular matrix components and adhesion-related signaling molecules serve important roles in the heart, as evidenced by human genetic disorders affecting the aorta and aortic valve, such as Marfan, Williams, Loeys-Dietz, and Weill-Marchesani syndromes. Here, we identify new functions of the multiple EGF-like domain protein ITGBL1 given its human aortic expression and integrin sequence homology. Our results indicate ITGBL1 is a highly-conserved protein expressed in multiple species and homologous to the beta integrin stalk region. Altering Itgbl1 levels consistently affected zebrafish ventricular size while itgbl1 RNA introduction ameliorated the abnormality, suggesting a role of itgbl1 in zebrafish cardiac development. To identify potential molecular functions, human ITGBL1 expression and cell-based studies were performed. Expression data analysis for potential ITGBL1-associated pathways demonstrated significant enrichment for the extracellular matrix, adhesion and collagen functional processes. We found that cells engineered to express ITGBL1 demonstrate selective differential adhesion to collagens. Furthermore, collagen-dependent vascular endothelial cell network formation was inhibited by ITGBL1, independent of cell death. ITGBL1-expressing cells decreased SMAD reporter activity, suggesting ITGBL1 may influence the TGF-beta pathway and play role in extracellular matrix-dependent signaling. To evaluate the potential influence of ITGBL1 on human outflow tract abnormalities, we assessed genomic structural variation in individuals with hypoplastic left heart for 13q33.1 variation. We show that ITGBL1 is copy number variable in a hypoplastic left heart individual, and resulting truncation of the protein structure yields a continuing ITGBL1 function on adhesion verified in cells. These data support a fundamental role for the multiple EGF-like protein ITGBL1 in the heart. To our knowledge, these results are the first description of ITGBL1 potential roles in development.

We have developed and genotyped >15,000 SNP assays by using a primer extension genotyping assay with fluorescence polarization (FP) detection. Although the 80% Success rate of this assay is similar to those of other SNP genotyping assays, we wanted to determine the reasons for the failures and find ways to improve the assay. We observed that the failed assays fell into three general patterns: PCR failure, excess of heterozygous genotypes, and loss of FP signal for one of the dye labels. After analyzing several hundred failed assays, we concluded that 5% of the assays had PCR failure and had to be redesigned. We also discovered that the other two categories of failures were due to misincorporation of one of the dye-terminators during the primer extension reaction as a result Of primer shortening with a reverse reaction involving inorganic pyrophosphate, and to the quenching of R110-terminator after its incorporation onto the SNP primer. The relatively slow incorporation of R110 acycloterminators by AcycloPol Compounds the problem with the R110 label. In this report, we describe the source of the problems and simple ways to correct these problems by adding pyrophosphatase, using quenching as part of the analysis, and replacing R110 by Texas red as one of the dye labels. With this new protocol, we have achieved similar to95% Success rate in assay development Without the need for optimization.