Schaid, Daniel J.

Abstract Several platforms for genome sequencing and methods for exon enrichment are now available and continue to evolve quickly. In this study, we compared the performance of whole exome sequencing using two instruments, Illumina's GAIIx and HiSeq 2000, comparing single versus duplicate runs. Eight runs of eight samples each were performed with genomic DNA derived from whole blood and sample libraries prepared with the Agilent SureSelect capture array. One run was conducted in duplicate on the GAIIx, and, for some analyses, the combined data were used. We used TREAT (Targeted RE-sequencing and Annotation Tool), developed in-house for data analysis, including sequence alignment (MAQ, BWA), local re-alignment (GATK), variant calling (utilizing MAQ), annotation (SIFT and Seattle Seq), and visualization. Overall, duplicate runs on the GAIIx platform (compared to a single run) increased the number of total reads by approximately two-fold: from ∼66 M to ∼130 M. For both single and duplicate runs, ∼90% of reads mapped to the reference sequence, and ∼55% of reads mapped on-target. Of importance, the percent coverage of the target region increased substantially with duplicate runs; ∼72% of the target region was covered at 40-fold or more when run in duplicate compared to only ∼52% when run once. At 30x, 20x, and 10x, there was a 1.3- (62% v. 78%), 1.2- (72% v. 84%), and 1.1-fold (84% v. 91%) increase respectively, in the percent coverage of the target region for samples run twice. A similar number of filtered, on-target SNPs per sample (∼24 K) was found for both single and duplicate run analyses. However, a 1.3-fold increase in the number of on-target indels was seen in the duplicate run (∼1,200) compared to the single run (∼900). When the combined data for samples run twice on a GAIIx were compared to the samples run using a HiSeq, the total number of reads were similar (∼130 M), although several of the samples had substantially more reads on the HiSeq platform. Samples run on a HiSeq had an increase in the percent coverage of the target region at 40x (72% v. 81%), 30x (78% v. 86%), 20x (84% v. 91%) and 10x (91% v. 95%) compared to the two-run approach on a GAIIx. A similar number of filtered on-target SNPs (∼24 K/sample) and indels (∼1200) was found with the HiSeq compared to two runs on a GAIIx. In summary, our results demonstrate increases in the total number of reads and in overall coverage with the HiSeq 2000 instrument. The total number of SNPs and indel that mapped on target for the version 1 Agilent SureSelect capture array is ∼25 K and 1,200 per sample, respectively. As capture is of variable efficiency for individual runs, higher average coverage of the target region is necessary for sufficient coverage of poorly captured regions. Additional analyses comparing Agilent SureSelect versions 1 and 2 arrays are underway. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 2742. doi:10.1158/1538-7445.AM2011-2742

Abstract Introduction: The NSABP P-1 and P-2 breast cancer prevention trials have demonstrated that the SERMs tamoxifen and raloxifene can reduce the risk of breast cancer in high risk women. Methods: We performed a nested matched case-control GWAS utilizing DNA from participants enrolled in P-1 and P-2. 592 participants developed invasive breast cancer or ductal carcinoma in situ, and were matched with 1171 controls who did not. Genome-wide genotyping with the Illumina 610-Quad chip and CYP2D6 genotyping for pertinent alleles were performed. Functional genomic studies involved the use of cultured cells, with siRNA knockdown, followed by qRT-PCR and Western blot analyses. Results: Eleven SNPs with p-values < 3E-05 were identified. Twenty one additional SNPs with p < 4E-05 were identified by imputation around the genotyped SNPs on chromosomes (Chrs) 4, 8, 9, 13, 16 and 22, which were then genotyped with the Invader platform. Initial functional genomic studies focused on 6 SNPs on Chr16 (p-values 1.81-9.55E-06), all of which were in ZNF423, a gene encoding a putative zinc-finger protein. The Chr16 minor variant SNP sequences had less risk than the common variant (OR=0.7). Our functional studies showed that incubation of Hs578T cells stably transfected with estrogen receptor (ER) α with 0.1 nM estradiol (E2) induced expression of ZNF423, BRCA1/2 for WT but not variant SNP sequences. There was also a striking difference between E2-induced ZNF423, BRCA1/2 expression for variant and WT SNPs in lymphoblastoid cell lines (LCLs) stably transfected with ERα, with only WT showing induction. ChIP assay showed that ZNF423 could bind to the 5’-flanking region of BRCA1, and reporter gene assays showed that ZNF423 could regulate BRCA1/2 transcription. In LCLs stably transfected with ERα and genotyped for ZNF423 SNPs, estrogen-dependent expression of all 3 genes occurred only in the presence of the WT, but not variant ZNF423 SNP sequences. Blockade of ERα with 4-OH tamoxifen prevented induction of expression in cells with WT sequences, but resulted in gene dose-dependent increases in BRCA1/2 expression in cells with variant SNP sequences-suggested that the minor variant has a “protective” effect during clinical SERM therapy and verified by ChIP assays. Conclusions: The SNPs in ZNF423 identified during GWAS were associated with differential E2-dependent induction of ZNF423, BRCA1 and BRCA2-implying that ZNF423 is “upstream” for the estrogen-dependent induction of BRCA1/2 expression. However, LCLs with variant SNP sequences showed greatly enhanced expression of ZNF423, BRCA1/2 during 4-OH tamoxifen exposure, while cells with WT sequences did not. These results may reveal a novel mechanism for SERM-dependent regulation of BRCA1/2 expression and may have implications for patient selection for SERM therapy. Funded by U10-CA37377, U10-CA69974 and U14-GM61388. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 4727. doi:10.1158/1538-7445.AM2011-4727

Abstract Background: Candidate gene and candidate pathway-based pharmacogenomic studies have been used to investigate the inherited variability of prognosis of lung cancer treated with platinum-based chemoptherapy. In order to make it possible to conduct an ‘unbiased’ query across the entire genome, a genome-wide association study (GWAS) in 283 lymphoblastoid cell lines (LCLs) was carried out to help identify candidate SNPs associated with variation in cisplatin cytotoxicity. Furthermore, we set out to test the hypothesis that genetic variation resulting in differences in cisplatin cytotoxicity might influence the overall survival of patients with lung cancer treated with platinum-based antineoplastic agents. Material and Methods: We applied ethnically diverse Coriell “Human Variation Panel” LCLs for GWAS. Then, we picked 168 top significant SNPs selected from GWAS in 283 LCLs and genotyped these SNPs in 1183 lung cancer patients (222 SCLC and 961 NSCLC) who received platinum-based chemotherapy. Genotyping was performed using the Illumina Golden Gate platform. The association of SNPs with lung cancer overall survival was analyzed by Cox regression model after adjusting for disease stage. Functional validation of candidate genes was performed by siRNA screening in NSCLC cell lines CRL5872, H1299 and SCLC cell line CRL5823. Results: A total of 157 SNPs were successfully genotyped for 1183 lung cancer patient samples. The Cox regression analysis indicated that 9 and 10 different SNPs were associated with overall survival for NSCLC and SCLC patients, respectively (p value <0.05). Fifteen genes including 7 genes located in the region of 200kb up or downstream of 4 SNPs (rs11169748, rs2440915, rs5952066 and rs7620841) and 8 genes for which the expression was correlated with three SNPs (rs11169748, rs2440915 and rs5952066) were selected for siRNA screening. Knockdown of DAPK3 and METTL6, for which the expression levels were correlated with the SNPs rs11169748 and rs2440915, significantly decreased cisplatin sensitivity in lung cancer cell lines CRL5872, H1299 and CRL5823. Conclusions: This series of clinical and complementary laboratory-based functional studies identified several candidate genes and SNPs that might help predict treatment outcomes of platinum-based therapy in lung cancer. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 1285. doi:10.1158/1538-7445.AM2011-1285

Abstract Introduction: Androstenedione is a precursor for estrogen biosynthesis catalyzed by aromatase. Plasma androstenedione concentrations in postmenopausal women with early stage breast cancer display large interindividual variation (Ingle et al., Cancer Res. 70:3278–86, 2010). We performed a GWAS to identify single nucleotide polymorphisms (SNPs) associated with plasma levels of androstenedione in 776 postmenopausal women with resected early stage breast cancer. Methods: Plasma androstenedione concentrations were measured by GC-MS/MS. GWAS was performed using Illumina Human610-Quad BeadChips. The GWAS results were controlled for population stratification (PS) by Eigenstrat analyses. Plasma androstenedione concentrations were associated with genotype by using a quasi-likelihood model with adjustment for PS, BMI, age and other clinical variables associated with androstenedione concentrations (p<0.01). Imputation using HapMap2 data was performed within 200 kb on either side of regions containing SNPs with p<E-05. Results: 563,945 SNPs were analyzed, followed by imputation. The genotyped SNP with the lowest p value, rs4736317 (p=4.61E-06), was on chromosome 8 near the CYP11B1 and CYP11B2 genes. This SNP had a multiplicative effect of 1.128 (95% Cl: 1.07, 1.19) and a MAF of 0.40. After imputation, 20 SNPs with apparent p<E-06 were observed in this same region, with the lowest p value for rs9773022 (p=8.1E-08). We then performed functional genomic studies, and observed that two of the SNPs with low p values (rs4736312, p=1.49E-05 and rs4736350, p=6.34E-06) created apparent Oct-1 transcription factor binding sites (TFBS). Differential nuclear protein binding of Oct-1 to probes for these SNPs was verified with EMS and supershift assays using nuclear extract from A549 and NCI-H295R cells. Treatment of these two cell lines with 10 nM androstenedione led to the disruption of Oct-1 binding with the A (wild type) allele on rs4736312. rs4736312 is located in the 3'-UTR of CYP11B1 15 bp distant from response elements for the progesterone receptor and the glucocorticoid receptor. We also demonstrated that CYP11B1 and CYP11B2 were induced by estradiol in U2OS cell lines that express ER receptor and by androstenedione in H295R cells that express the androgen receptor. Transfection of a reporter gene construct with the rs4736312 SNP sequences into COS-1 and H295R cells suppressed reporter gene activity for the variant SNP sequence as compared to wild type, an observation that may correlate with the higher plasma levels of androstenedione in women (549× 328 vs 410×208 mean×SD, M/ml) with the variant allele. Conclusion: We have identified a series of SNPs in a region of chromosome 8 in and around CYP11B1 and CYP11B2 that are associated with elevated plasma androstenedione concentrations in postmenopausal women with resected early stage breast cancer. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2011 Nov 12-16; San Francisco, CA. Philadelphia (PA): AACR; Mol Cancer Ther 2011;10(11 Suppl):Abstract nr C114.

Current genome‐wide association studies still heavily rely on a single‐marker strategy, in which each single nucleotide polymorphism (SNP) is tested individually for association with a phenotype. Although methods and software packages that consider multimarker models have become available, they have been slow to become widely adopted and their efficacy in real data analysis is often questioned. Based on conducting extensive simulations, here we endeavor to provide more insights into the performance of simple multimarker association tests as compared to single‐marker tests. The results reveal the power advantage as well as disadvantage of the two‐ vs. the single‐marker test. Power differentials depend on the correlation structure among tag SNPs, as well as that between tag SNPs and causal variants. A two‐marker test has relatively better performance than single‐marker tests when the correlation of the two adjacent markers is high. However, using HapMap data, two‐marker tests tended to have a greater chance of being less powerful than single‐marker tests, due to constraints on the number of actual possible haplotypes in the HapMap data. Yet, the average power difference was small whenever the one‐marker test is more powerful, while there were many situations where the two‐marker test can be much more powerful. These findings can be useful to guide analyses of future studies. Genet. Epidemiol. 36:480‐487, 2012. © 2012 Wiley Periodicals, Inc.

AbstractGene‐set analyses have been widely used in gene expression studies, and some of the developed methods have been extended to genome wide association studies (GWAS). Yet, complications due to linkage disequilibrium (LD) among single nucleotide polymorphisms (SNPs), and variable numbers of SNPs per gene and genes per gene‐set, have plagued current approaches, often leading to ad hoc “fixes.” To overcome some of the current limitations, we developed a general approach to scan GWAS SNP data for both gene‐level and gene‐set analyses, building on score statistics for generalized linear models, and taking advantage of the directed acyclic graph structure of the gene ontology when creating gene‐sets. However, other types of gene‐set structures can be used, such as the popular Kyoto Encyclopedia of Genes and Genomes (KEGG). Our approach combines SNPs into genes, and genes into gene‐sets, but assures that positive and negative effects of genes on a trait do not cancel. To control for multiple testing of many gene‐sets, we use an efficient computational strategy that accounts for LD and provides accurate step‐down adjustedP‐values for each gene‐set. Application of our methods to two different GWAS provide guidance on the potential strengths and weaknesses of our proposed gene‐set analyses.

AbstractOver the past several years, genome‐wide association studies (GWAS) have succeeded in identifying hundreds of genetic markers associated with common diseases. However, most of these markers confer relatively small increments of risk and explain only a small proportion of familial clustering. To identify obstacles to future progress in genetic epidemiology research and provide recommendations to NIH for overcoming these barriers, the National Cancer Institute sponsored a workshop entitled “Next Generation Analytic Tools for Large‐Scale Genetic Epidemiology Studies of Complex Diseases” on September 15–16, 2010. The goal of the workshop was to facilitate discussions on (1) statistical strategies and methods to efficiently identify genetic and environmental factors contributing to the risk of complex disease; and (2) how to develop, apply, and evaluate these strategies for the design, analysis, and interpretation of large‐scale complex disease association studies in order to guide NIH in setting the future agenda in this area of research. The workshop was organized as a series of short presentations covering scientific (gene‐gene and gene‐environment interaction, complex phenotypes, and rare variants and next generation sequencing) and methodological (simulation modeling and computational resources and data management) topic areas. Specific needs to advance the field were identified during each session and are summarized.Genet. Epidemiol. 36 : 22–35, 2012. © 2011 Wiley Periodicals, Inc.