Christopher A. Haiman - US grants

DESCRIPTION (provided by applicant): This is a resubmission of a competing renewal application requesting support for a Program Project currently in its 27th year of continuous funding. At the recommendation of the Program's External Advisory committee, the theme of this renewal application was narrowed from the previous broad theme of "latrogenic Causes of Cancer" to a more narrow theme of "Unresolved Public Health Issues Related to HT and Cancer." The Program has a rich history of research on exogenous hormones including hormone therapy (HT). Data generated from the Program have had major influences on pharmaceutical formulations, prescribing practices, and chemoprevention strategies regarding exogenous hormone especially HT use. The scientific program of the current application consists of four projects focused around this theme supported by four core resources. The projects include: (1) a study of HT and ovarian cancer emphasizing possible risk differences in formulations and histologic subtypes; (2) a study of HT and non-Hodgkin's lymphomas (NHL), again with an emphasis on possible risk differences by histologic subtypes; (3) a study to determine which women who use HT develop adverse changes in mammographic densities and whether these are the women, in turn, who will develop breast cancer; and (4) a statistical study of the impact of different sources of measurement error in studies of HT and cancer. As a major emphasis of the Program will be to evaluate the potential risk modifying effects of a series of genes contributing to absorption, transport and metabolism of HT, this latter project also deals with statistical issues related to identification and analysis of reconstructed haplotypes. Core resources include a Scientific Leadership and Administrative Core (Core A) for overseeing scientific direction and conducting the administrative activities required by this Program; the Cancer Surveillance Core (Core B), which provides case identification, including especially rapid case ascertainment for the projects; a Genotyping Core (Core C), for genotyping a large series of candidate genes involved in HT absorption, transport and metabolism and selected additional genes; and a Control Identification Core (Core D), which provides neighborhood control identification for two of the proposed projects.

[unreadable] DESCRIPTION (provided by applicant): Excess exposure to endogenous and exogenous sources of estrogen heightens cell division in the endometrium which is thought to increase the rate at which somatic mutations occur and become fixed in subsequent mitoses, and serve as initiating events in the development of cancer. Multiple highly conserved mechanisms have evolved to identify and cope with constant insults on DNA from endogenous and exogenous sources and to maintain genomic integrity. Many human cancer syndromes are caused by rare germline mutations in DNA repair genes that result in deficiencies in normal DNA repair capacity. One such syndrome, caused by mutations in mismatch repair genes, is hereditary nonpolyposis colorectal cancer. Endometrial cancer is a common phenotype of this familial syndrome, directly implicating DNA repair capacity in the pathogenesis of this disease. To date, the role of low-penetrant alleles in genes involved in mismatch repair or other DNA damage repair or response pathways has not been thoroughly studied in relation to endometrial cancer risk. This current project builds on existing resources to study efficiently and comprehensively, the importance of common genetic variation in 61 candidate genes in DNA repair-related pathways on endometrial cancer risk. Over the past year we have constructed an optimal array of "tagging" SNPs designed using Illumina genotyping technology that provide high predictability of common genetic variation in these candidate genes across multiple racial-ethnic populations. In this application, we will utilize prospectively collected biospecimens, risk factor and outcome data from two established endometrial cancer case-control studies nested within the Multiethnic Cohort study (cases, n = 542; controls, n = 1,234) and the California Teachers Study (cases, n = 580; controls, n = 1,160) to assess associations between these selected tagging SNPs (approximately 1,200) and all known missense variants (approximately 300) in these genes in relation to endometrial cancer risk. This study will use novel high-throughput genotyping technology as well as rigorous statistical analytic methods to account for multiple hypothesis testing, and allows for replication and cross- validation of findings across these cohorts. The ability to accurately identify high-risk sub-groups may have profound implications on endometrial cancer prevention as well as advance research aimed at targeted drug development. [unreadable] [unreadable] [unreadable]

DESCRIPTION (provided by applicant): Prostate cancer is a leading cause of morbidity and mortality among men. Risk factors for prostate cancer have remained elusive and aside from age, having a family history of the disease or African ancestry, until recently, no genetic or non-genetic (i.e. lifestyle) risk factors have been consistently demonstrated to contribute to variation in disease risk in the population. Over the past year, we contributed to the first reproducible genetic risk factor for prostate cancer at 8q24. Through fine-mapping of 8q24 in five racial/ethnic populations in the Multiethnic Cohort Study (MEC), we identified three regions that contain 7 independent risk variants for prostate cancer. Only some of these variants were found in studies conducted in European Whites, thus, emphasizing the power of genetic studies in multiple racial/ethnic populations to reveal a more complete spectrum of variants associated with disease risk. In this application, we propose to work with the GEI-GWA Steering Committee to identify genetic factors that contribute to prostate cancer by performing a well-powered genome-wide association study among African American, Japanese American, Native Hawaiian, Latino and European American men in the MEC. More specifically, we propose the genotyping of approximately 1,000,000 single nucleotide polymorphisms (SNPs) and >940,000 probes to detect copy number variation in 3,750 prostate cancer cases and 3,750 controls, to identify pan-ethnic genetic variants that affect risk in all five racial/ethnic populations, as well as important ethnic-specific variation. In this multi-ethnic dataset, we will also examine interaction between associated variants, environmental factors (thereby better defining the role of these factors) and disease severity, as well as explore the causes of heterogeneity of genetic effects across populations. We have also established collaborations with other prostate cancer researches to replicate the ethnic-specific associations observed in this GWA study in minority populations. We expect this work to significantly advance knowledge of the etiology of prostate cancer across these racial/ethnic populations, guiding the development of future preventive, early detection, prognostic and even therapeutic measures.

DESCRIPTION (provided by applicant): The 8q24 locus harbors common susceptibility alleles for prostate cancer. Through fine-mapping of 8q24 in five racial-ethnic populations in the Multiethnic Cohort Study (MEC), we identified three regions that contain 7 independent risk variants for prostate cancer. Only some of these variants were found in studies conducted in European Whites, thus, emphasizing the power of fine-mapping in multiple racial-ethnic populations to reveal a more complete spectrum of variants associated with disease risk. Further highlighting the breakthrough nature of the discoveries at the 8q24 locus this region has even more recently been found to harbor common susceptibility alleles for breast cancer and colorectal cancer (CRC). Genome-wide association (GWA) studies of breast and CRC are underway among populations of European ancestry and a number of other risk loci have been identified and confirmed. In this application, we propose to fine-map the 8q24 locus and the genomic regions identified in these breast and CRC GWA studies in multiple racial-ethnic populations with diverse genetic backgrounds and environmental exposures, to more comprehensively enumerate and localize the risk variants in these regions and to assess genetic risk heterogeneity in the population. In all regions, we will evaluate SNPs selected to highly predict all common alleles as defined in the HapMap populations. We will also perform exon resequencing of genes in candidate regions identified in GWA studies to test common coding variants. This work will be conducted in large case-control studies comprised of African Americans, Japanese Americans, Native Hawaiians, Latinos, and European Americans from the MEC that include 2,650 breast cancer cases and 2,900 controls, and 1,800 colorectal cancer cases, and 4,500 controls. Heterogeneity of genetic effects within and across populations, resulting from differences in linkage disequilibrium as well as in genetic, environmental and lifestyle risk factors, will be assessed to better define the susceptible subgroups of the population. We expect this work to significantly advance knowledge of the etiology of breast and CRC across these racial-ethnic populations, and to reveal a more complete picture as to the contribution of germline variation in these regions to breast and CRC risk;work that we hope will guide the development of future preventive, early detection and prognostic strategies. PUBLIC HEALTH RELEVANCE: The goal of this project is to identify common risk alleles for breast and colorectal cancer in African American, Japanese American, Native Hawaiian, Latino and European American populations in the prospective Multiethnic Cohort Study. More specifically, we propose to fine-map genetic variation at chromosome 8q24 as well as regions identified in genome-wide scans in European Whites, in multiethnic samples, to localize functional variants as well as reveal the full spectrum of common alleles that contribute to risk of breast and colorectal cancer in the general population. We will also incorporate environment and lifestyle risk factors in association testing to assist in resolving potential heterogeneous genetic effects across populations, and, better define the subgroups of the population at greatest risk of developing these common cancers. We expect findings from this work to guide the development of future preventive, early detection and prognostic strategies

DESCRIPTION (Provided by the applicant): Prostate cancer is one of the leading causes of morbidity and mortality among men. At the moment, we do not understand the underlying etiology, or why different groups, such as African Americans, have higher rates of disease. Risk factors for prostate cancer have remained elusive and aside from age, having a family history of the disease or African ancestry, until recently, no genetic or non-genetic (i.e., lifestyle) risk factors have been consistently demonstrated to contribute to variation in disease risk in the population. In this application, we propose to undertake a large-scale collaborative effort to uncover genetic predictors of prostate cancer in African American men. For this effort, we have assembled a multi-institutional team of investigators with experience in prostate cancer research in minority populations who are eager and willing to pool resources, specimens and data from their established studies, and to work closely together towards a common goal. To identify genetic factors that contribute to prostate cancer risk in African American men we propose to conduct a well-powered multi-stage genome-wide association study. We currently have funding (U01-NCI) to genotype 1,000,000 single nucleotide polymorphisms (SNPs) in 2,500 African American prostate cancer cases and 2,500 controls. We now propose to expand the size of stage 1 to 4,086 cases and 4,145 controls. In stage 2 we will perform follow-up genotyping of 1,500 SNPs that demonstrate significant main effects and interactions with known risk variants for prostate cancer (e.g. 8q24) in an additional 1,044 African American cases and 1,083 controls. In this study we will also assess the pan-ethnic effects of the variants identified in the African American scan, by replication testing in 600 prostate cancer cases and 1,050 controls from West Africa. In this dataset, we will also examine interactions between associated variants, environmental factors (thereby better defining the role of these factors) and disease severity. We expect this work to significantly advance knowledge of the etiology of prostate cancer and racial/ethnic disparities in prostate cancer risk, and to guide the development of future preventive, early detection, prognostic and even therapeutic measures. PUBLIC HEALTH RELEVANCE: The goal of this project is to identify common risk alleles for prostate cancer in African American men. For this effort, we have assembled a multi-institutional team of investigators with experience in prostate cancer research in minority populations who are eager and willing to pool resources, specimens and data from their established studies, and to work closely together towards a common goal. More specifically, we propose to conduct a multi-stage genome-wide association study of prostate cancer, which will include >5,500 African American cases and >5,100 African American male controls, to better define the subgroups of the African American population at greatest risk of developing this common cancer.

DESCRIPTION (provided by applicant): We propose an integrated multiple-PI Project to systematically discover and replicate additional common genetic variants associated with breast cancer, assess their biological significance, and develop evidence based assessments of the clinical validity of prediction algorithms using these variants, and their suitability for translation into clinical practice. In sub-Project 1 we will combine the resources of (a) major GWAS for breast cancer amounting to >15,000 cases and (b) three pre-existing Consortia with over 48,000 additional cases to provide the large sample size needs necessary in the replication phase of GWAS. We will fine map the associated loci in collaboration with the major Consortia conducting GWAS for breast cancer in Asian and African-American women. In sub-Project 2 we will conduct a series of investigations to (a) assign a gene function to each replicated risk variant by measuring expression of 24,000 RNA transcripts in breast tumor tissue and normal tissue, from women for whom we also have an lllumina 540 GWAS available; by identifying networks of genes in which alterations of expression can be linked to specific germline risk variants; and by using Chromosomal Conformation Capture assays to examine whether associated intergenic regions fold physically in a way that brings them into contact with distant genie regions. We will also (b) examine whether loss or gain of function of the genes implicated in (a) in breast epithelial cells or stromal cells alter phenotypes in vitro in a 3-D model of breast morphogenesis and oncogenesis. In sub- Project 3, we will develop breast cancer prediction models that can be used to stratify women according to breast cancer risk. We will attempt to discover gene-gene interactions by reanalyzing the GWAS data, and we will systematically examine the genome-wide significant gene variants for effect modification by established breast cancer risk factors, using the largest set of prospective studies available. We will develop and refine risk models that incorporate both the germline risk factors and the established non-genetic risk factors, and also assess these in a cohort of women with higher familial risk of breast cancer (to specifically address the clinical needs of women at high risk due to a strong family history of breast cancer). Finally, we will analyze data from the major trials of primary prevention of breast cancer to address the question of whether the protective effect of tamoxifen is altered by risk status for our prediction models, data with a direct bearing on clinical decision-making with respect to chemoprevention for women at known high risk.

DESCRIPTION (provided by applicant): We propose to establish a multidisciplinary team of epidemiologists, statisticians, basic scientists and clinician scientists to promote and pursue thorough identification and characterization of genomic loci associated with prostate cancer. ELLIPSE (ELucidating Loci Involved in Prostate cancer SuscEptibility) is a translational grant comprised of three highly integrated projects. Project 1 aims to take advantage of existing and shortly to be completed GWAS of prostate cancer in European, African American, Latino and Japanese populations to discover novel risk loci, and to rigorously replicate these associations in large existing consortia of prostate cases and controls (PRACTICAL, BPC3, and MADCaP). A major focus of this expanded effort is to identify loci that may selectively be associated with advanced disease and variants that contribute to ethnic difference in disease risk. Selective fine mapping of risk loci will also be performed to comprehensively characterize the relevant allelic locus, utilizing all available data on common variants from Hapmap, selective sequencing efforts, and from the 1000 Genome Project. Project 2 is focused on understanding the gene(s) that the non-protein coding risk variants are acting through. Two hypotheses will be systematically explored using a wide variety of established and emerging techniques: the risk loci harbor as yet undetected transcripts (either coding or non-coding) and the risk loci are regulatory elements. Finally, in Project 3 we propose to investigate the potential of this new knowledge on the genetic basis of prostate cancer susceptibility to enhance risk assessment, through gene-gene and gene-environment interactions, and importantly, to provide the potential for novel clinical practices through impacts on cancer diagnosis and treatment, or newer cancer prevention strategies. The overarching goal is to discover the pathways that drive prostate cancer pathogenesis and to assess their role in clinical decision making.

DESCRIPTION (provided by applicant): Multiple myeloma (MM) is a malignancy consisting of a neoplastic clone of plasma cells, the terminally-differentiated B lymphocytes responsible for antibody production. It is the second most common cause of death among the hematological malignancies. Despite recent advances in treatment, the 5-year survival rate remains under 40%. African Americans experience incidence rates over 2-fold higher than those of Whites;the reasons for the racial differences in disease rates are unknown. Risk is also 2- to 3-fold elevated among persons with a family history of multiple myeloma and/or its precursor, monoclonal gammopathy of undetermined significance (MGUS). We hypothesize that there is a genetic component to multiple myeloma risk and that part of the excess risk in African Americans is mediated through inherited variation. In this application, we propose to perform a genome-wide association study to identify common genetic variants that contribute to multiple myeloma risk in the African American population. More specifically, we will examine 1,000,000 genetic markers in 2,072 African American multiple myeloma cases and 4,645 African American controls. Cases for this study will be recruited and specimens collected by means of a network of collaborations with NCI/SEER population-based cancer registries (California, Detroit, Louisiana and New Jersey) and clinical centers in 6 regions with large African-American populations (University of Alabama at Birmingham, Johns Hopkins University, Emory University, Karmanos Cancer Institute/Wayne State University, Northwestern University, University of Chicago, and M.D. Anderson Cancer Center/University of Texas). Controls will consist of 4,645 healthy unaffected African-American subjects from across the U.S. We will assess the generalizabilty of our findings by association testing of the top 100 SNPs in 1,500 MM cases and 3,710 controls of European origin in multiple existing studies from the U.S. and Australia. In addition, with NCI consultant Dr. Landgren, we will examine the associations of the risk variants with the precursor, MGUS among 700 African-American subjects with MGUS and 1,400 controls recruited from NHANES. In this proposal we will utilize the strengths of collaborative, large-scale genetic research to identify common risk alleles for MM in a population (African American) that experiences a health disparity with respect to this cancer. In addition to illuminating etiology of multiple myeloma in African Americans, we hope to discover new pathways that can be exploited for drug development. Our ultimate goal is to reduce the health disparity experienced by African Americans due to this fatal cancer. PUBLIC HEALTH RELEVANCE: Multiple myeloma (MM) is a cancer of blood cells (lymphocytes) with a 5-year survival rate of under 40%. This cancer is important because it is highly fatal and because it disproportionately affects an underserved population. For unknown reasons, African Americans have among the highest risks of MM in the world and are affected more than twice as often as White Americans. In this proposal we will collaborate with other institutions across the U.S. to collect samples from 2,072 African American MM patients and compare their genetic variation to that of 4,645 African American controls (with previously genotyped DNA). In addition to illuminating the cause of multiple myeloma in African Americans, we hope to discover new pathways that can be used for drug development. Our ultimate goal is to reduce the health disparity experienced by African Americans due to this fatal cancer.

DESCRIPTION (provided by applicant): Known genetic risk factors for breast cancer account for only ~30% of the familial risk of the disease (so-called 'missing heritability') with common variants (frequency >10%) revealed through genome-wide association studies (GWAS) explaining one-third of this percentage. A large fraction of familial risk is likely due to variants that are less common (1-10%) or rare (<1%); a space of genetic variation that has yet to comprehensively explored in relationship with breast cancer risk. In this application, we propose to undertake a large-scale collaborative effort to uncover genetic predictors of breast cancer in women at high familial/genetic risk. For this effort, we have assembled an international team of investigators with experience in breast cancer research who are eager and willing to pool resources, specimens and data from their established studies, to search for novel and less common risk variants for this major cancer. In Aim 1, we propose to conduct a well-powered genome-wide association study (with 80% power to detect a relative risk of 1.5 or more, or 0.67 or less, for a variant with frequency as low as 1%). In stage 1, we will genotype 5 million SNPs for 3,000 breast cancer cases at increased familial/genetic risk, based on having a strong family history of the disease, and 3,000 controls of European ancestry. In stage 2, we will follow-up the 500 most significant associations using an additional 17,000 breast cancer cases and 17,000 controls of European ancestry. Novel validated risk variants will be examined in African American, Latino and Japanese samples, as well as in relationship with breast cancer tumor subtypes. A second Aim of this study will be to conduct a hypothesis generating GWAS analysis of estrogen receptor positive and estrogen receptor negative breast cancer in women at high familial/genetic risk in search of risk variants that are specific for these tumor subtypes. We will also estimate the amount of familial aggregation (polygenic variance; heritability) explained by all known risk variants (Aim 3), including those discovered in Aim 1, using population-based case family studies with detailed information about family history and DNAs from relatives from the Breast Cancer Family Registry (BCFR). Our goal is to improve upon the comprehensive risk model BOADICEA for estimating a woman's lifetime risk of breast cancer based on her genetic, family history and epidemiologic profile. We expect this work to significantly advance knowledge of the etiology of breast cancer and to guide the development of future preventive, early detection, prognostic and even therapeutic measures that will have wide clinical and public health utility.

DESCRIPTION (provided by applicant): African American men have two times or more the incidence rate of prostate cancer than other U.S. populations, and high rates of prostate cancer are also seen in other African and African-derived populations. It is generally accepted that both common and rare genetic variants contribute to risk of complex diseases such as prostate cancer, however their relative contributions to overall heritability is a subject of intense controversy. For rare variants to have a significant influence on the risk of complex disease the spectrum of effect sizes must be considerably larger in magnitude than for common variants; to date however there is only limited evidence for or against this hypothesis since the means of comprehensively testing rare variation in the genome has not been possible until very recently. In this proposal we seek to test the contributions of both common and rare genetic variants to the risk of prostate cancer in men of American ancestry using a targeted genome-wide association study approach. In Aim 1 we plan to sequence (at 30x coverage) the exome and regulatory regions (~160 Mb) of the genome, as defined by epigenetic marks in prostate cancer cell lines, in 1,000 men of African ancestry (500 with aggressive disease) and 1,000 controls. Both single SNP and burden of rare variants analyses will be performed and replication testing of the most statistically significant sequence variations (~24,000) will be examined in additional samples of African ancestry (6,000 cases and 6,000 controls) through the African Ancestry Prostate Cancer (AAPC) consortium. In addition to association testing of single variants, we will conduct burden of rare variants analyses of coding and non-coding variants at the gene and pathway level. In Aim 2, we will examine interactions between associated variants, environmental factors (thereby better defining the role of these factors) and disease severity. In Aim 3, we will assess biological function of the novel risk alleles identified in Aim 1 using a staged approach of eQTL analysis followed by in vitro analyses of enhancer activity as well as allele-specific effects. This proposal spans the spectrum of genetic epidemiologic research in prostate cancer, from genetic discovery (for all prostate cancer as well as aggressive disease) to gene environment interaction testing, to biological understanding. We expect this work to significantly advance knowledge of the etiology of prostate cancer and racial/ethnic disparities in prostate cancer risk, and to guide the development of future preventive, early detection, prognostic and even therapeutic measures.

We propose to utilize the Multiethnic Cohort (MEC), a PAGE Phase 1 study, to sample and assess genomic variation in well-phenotyped individuals of non-European (EA) ancestry. The MEC is a prospective population-based cohort of over 215,000 men and women in Hawaii and California that includes large representations of older adults for five racial/ethnic groups (African Americans, Latinos, Japanese Americans, Native Hawaiians and European Americans) at varying risks of chronic diseases. In the MEC, we have established a biorepository of blood and urine samples (N>74,000) linked to prospectively collected risk factor, biomarker and clinical/endpoint data. In PAGE2, we propose to explore the associations of DNA sequence variation for a broad range of phenotypes, focusing on conditions with disproportionate disease burdens in non-EAs. We propose to utilize the most comprehensive and cost-effective SNP genotyping array to capture and perform association tests in non-EA populations for: a) common and rare putative functional coding variation, and; b) predicted functional variants in ENCODE-defined regulatory elements in non-coding sequence at GWAS-identified risk loci. We will focus on diseases/traits/biomarkers that are common to PAGE2 studies and for which we have DNA available for large numbers of MEC subjects (e.g., type 2 diabetes, obesity, common cancers, fasting insulin, glucose, lipids, etc.). We will analyze single variants, risk scores, and burden of rare variants for single genes and pathways using both directional and omnibus methods. Using the comprehensive risk factor information available across PAGE2, we will also examine interactions between gene (e.g., SNP, risk score, burden), pathway and environment and build risk prediction models for the purpose of better defining the genetic contributions to racial/ethnic disparities in risk for common chronic diseases and traits. Last, we also propose to develop novel statistical methods and approaches for analyzing rare variants including: (a) an examination of kernel machine tests in the presence of subtle hidden population stratification; (b) extending the kernel machine methods to form tests of gene/pathway x environment and gene/pathway x gene/pathway interactions, and; (c) how to apply these methods to further characterize whether variants in specific pathways and their interactions explain racial/ethnic differences in risk. Working with other PAGE2 studies, the Coordinating Center and NHGRI, we hope to form a valuable population-based resource for the scientific community that will expand our understanding of ancestral differences in genomic disease associations and the role of genetic variation in understanding racial and ethnic differences in disease risk.

DESCRIPTION (provided by applicant): Obesity is a leading risk factor for metabolic and cardiovascular diseases and its prevalence has more than doubled since the 1980's, with the greatest burden carried by minority populations. Large-scale genome-wide associations studies (GWAS) have identified >70 genetic loci that are unequivocally associated with obesity-related traits primarily in European descent populations. So far, no large-scale GWAS for any obesity-related traits have been performed in Hispanic /Latinos (HL) populations, despite their increased prevalence of obesity. Although classified under one 'ethnic label', HL populations are incredibly diverse and genetically highly admixed with recent origins from Europe, Africa and the Americas. Hence, genome-wide association will necessitate a large collaborative effort and the use of advanced statistical methods (that go far beyond standard GWAS analyses) to account for and leverage their high degree of genetic diversity. Here, we propose to perform the first large-scale genomic study in search of obesity-susceptibility loci in HL populations. For aim 1, we have assembled the world's GWAS studies in HL populations, including >50,000 HL men and women with high-density SNP array data. Genome-wide imputation to multiethnic reference panels from the 1000 Genomes Project and other unique Amerindian resources will allow for comprehensive testing of common and low frequency variation present in HL populations as well as provide a rich resource for addressing genetic risk heterogeneity at obesity-related loci across HL sub-populations. To elucidate racial/ethnic transferability and fine-map association signals in aim 2, we will leverage data from large-scale GWAS of obesity-related traits in AA and EA populations that are available to us through our work with AA (n>50,000) and EA (GIANT consortium, n>200,000) consortia. In aim 3, we will employ functional analyses in Drosophila and bioinformatic data-mining tools to identify and characterize the target genes and functional alleles, and link associations with biological pathways. We are uniquely positioned and experienced to establish a large-scale collaboration to study the genomics of obesity in HLs. Our proposal is also unique and innovative for taking a GWAS study to the next translational stage, with an experimental research aim for further characterization of obesity specific genetic effects. Our study may improve the understanding of the genomic etiology of obesity, knowledge which may be used to reduce the burden of disease in underserved and understudied minority populations.

DESCRIPTION (provided by applicant): This application seeks support for the infrastructure of the Multiethnic Cohort (MEC) Study, which was established in Hawaii and southern California between 1993 and 1996 to study risk factors for cancer and other chronic diseases. The study was designed to take advantage of the ethnic and cultural diversity of the two geographic areas, as well as the expertise of the senior investigators in nutrition, ethnic/racial studies, and, subsequently, genetics. It is the most ethnically heterogeneous cancer cohort in existence. At baseline, the cohort included information on 215,000 men and women, comprised, by design, almost entirely of five ethnic/racial populations: Caucasians, Japanese Americans, Native Hawaiians, African Americans, and Latinos. The resource was later expanded to include a prospective bio repository of blood and urine specimens from ~ 70,000 of the participants. Leadership of the MEC entails a highly interactive, team approach; and the investigators have amply demonstrated their willingness to share data and participate actively in consortium projects. This application describes our aims over the next five years for maintaining and enhancing the infrastructure of the MEC, as well as plans for methodological research in the areas of genetic and nutritional epidemiology that utilize the resources of the cohort. Research accomplishments to date include significant contributions to understanding both genetic and environmental risk factors for cancer, particularly related to breast, prostate, colorectal and lun cancers. Nearly 250 papers describing these findings have been published. In addition, primarily over the last 20 years, more than 50 research grants have been built around the MEC, and more than 50 students and postdoctoral fellows have been trained on the study. This new grant will make possible the continuation of a well-integrated program of research aimed at evaluating environmental factors and genetic variants as risk factors for cancer and other common chronic diseases.

? DESCRIPTION (provided by applicant): Breast cancer is the most commonly diagnosed malignancy in the United States. The age-adjusted mortality rate of this cancer is more than 40% higher in African Americans (AAs) than in whites for reasons poorly understood. Since 2007, genome-wide association studies (GWAS) conducted in Asian and European descendants have identified nearly 100 susceptibility loci for this cancer. However, only a few of the initially identified risk variants can be directly replicated in AAs due to a small sample sizein previous studies and racial differences in genetic architectures and genetic/environmental modifiers. GWAS are often not equipped to study structural variants and are inefficient for capturing low-frequency variants. These variants, although virtually uninvestigated to date, are believed to contribute substantially to the heritability of breast cancer and other complex traits, particularly in African-ancestry populations. Furthermore, compared with Asian- and European-ancestry populations, the African-ancestry genome is much more heterogeneous and thus more informative, particularly as we expand the scope of genetic studies from common to less-common variants using next-generation sequencing technology. Herein, we propose a large consortium study in AAs to systematically search the whole genome to discover novel genetic susceptibility factors for breast cancer and further evaluate the influence of germline risk variants on breast cancer biology. Nearly 20,000 AA breast cancer patients and an equal number of controls will be included in this study. In Stage 1, we propose to sequence the whole genome for 1,200 breast cancer cases and 600 controls for association analyses. We will then use these sequencing data, along with data from other sources, to build a novel, comprehensive reference panel for imputation and meta-analysis of approximately 6,300 cases and 6,300 controls genotyped in four previous GWAS conducted in African-ancestry populations. We will utilize publically available genetic data, including functional genomic data, to enhance the abilit of the two aforementioned analyses to identify promising breast cancer susceptibility genes and variants for replication. In Stage 2, we will replicate approximately 60,000 promising variants in 5,500 cases and 5,500 controls. Genes/variants which show a promising association in Stage 2 will be evaluated further in Stage 3, including two additional stages (3A and 3B) in approximately 7,500 cases and 7,500 controls. Finally, we will use gene expression signatures to evaluate how germline risk variants identified in this study and previous studies affect the major signaling pathways of breast cancer. This proposed study will generate critically needed data in AAs to improve the understanding of the genetics, biology, and etiology of breast cancer.

? DESCRIPTION (provided by applicant): Prostate cancer (PCa) is the most common cancer in men with 220,000 new cases and 27,000 deaths estimated this year in the U.S. The vast majority of these deaths occur among the approximately 10-15% of cases diagnosed with aggressive PCa. There are few known risk factors for PCa beyond age, African descent and a family history of PCa, and there are no risk factors that can determine which men will develop aggressive versus non-aggressive disease. Multiple lines of evidence indicate a substantial heritable component of aggressive PCa. Over the past decade, genome-wide association studies (GWAS) have identified over 100 common susceptibility loci that collectively account for 33% of the familial risk of PCa. These loci contribute equally to risk of aggressive and non-aggressive disease which suggests they play a role in the very early stages of PCa tumor evolution. Recent sequencing studies have revealed rare coding variants (well under 1%) in genes such as BRCA1/2 and other DNA repair pathway genes that convey larger risks (3-5 fold) of aggressive PCa relative to non-aggressive disease. These observations suggest that the allelic architecture of aggressive disease may be quite different than overall PCa, and highlight the need for larger efforts focused on rare genetic variation (<1%). This spectrum of variation represents ~80% of all genetic variation in the human genome and is not adequately surveyed through GWAS. In this study, we will apply a multi-staged approach to reveal genes harboring rare variants that are associated with aggressive PCa. Whole-exome sequencing (Aim 1a) of 2,000 aggressive cases and 2,000 non-aggressive cases of European ancestry will be conducted followed by rare variant analysis of single sites and gene burden testing to identify novel susceptibility loci/genes for aggressive disease. We will validate the most significantly associated genes (~500) through targeted sequencing in an additional 7,500 aggressive and 7,500 non-aggressive cases (Aim 1b). Next, we will investigate the clinical predictive utility of the genes/variants identified in 2,300 cases in the STHM3 trial who are undergoing biopsy based on PSA and genetic risk score stratification (Aim 2). Last, we will examine whether the genes identified in Aim 1 contribute to the greater risk of aggressive PCa in 4,000 men of African ancestry (Aim 3). Through this tiered approach we expect to significantly advance knowledge of aggressive PCa etiology and health disparities as well as guide the development of early detection and prognostic strategies for the subset of men who are most susceptible to this fatal form of disease.

This renewal application seeks support for the infrastructure of the Multiethnic Cohort (MEC) Study, which was established in Hawaii and southern California in 1993-1996 to study risk factors for cancer and other chronic diseases. The study was designed to take advantage of the ethnic and cultural diversity of the two geographic areas, as well as the expertise of the investigators in nutrition, ethnic/racial studies, and genetics. It is the most ethnically diverse cancer cohort in existence. It achieves high cost-efficiency by supplementing active follow-up information with computerized linkages to SEER, vital statistics, hospital-discharge diagnoses, medical claim data, electronic medical records and geospatial information. At baseline, the cohort included information on 215,000 men and women, comprised almost entirely of five ethnic/racial populations: Caucasians, Japanese Americans, Native Hawaiians, African Americans, and Latinos. The resource was later expanded to include a prospective biorepository of blood and urine specimens from ~70,000 of the participants. Leadership of the MEC entails a highly interactive, team approach; and the investigators have amply demonstrated their willingness to share data/samples, and to participate actively in consortia. This application describes our aims over the next 5 years for maintaining and enhancing the infrastructure of the MEC, as well as plans for methodological research. Research accomplishments include significant contributions to understanding both genetic and environmental risk factors for cancer. Over 280 papers describing our findings have been published during the current grant cycle. In addition, over the last 20 years, 98 research grants have been built around the MEC, and more than 75 students and postdoctoral fellows have been trained on the study. This grant renewal will make possible the continuation of a well-integrated program of research aimed at evaluating environmental and genetic risk factors for cancer and other common chronic diseases, taking advantage of new approaches, such as genomics, microbiomics, adductomics and metabolomics. The MEC will allow the testing of innovative research hypotheses aimed at ensuring that progress in prevention applies to major US ethnic/racial minorities.

PROJECT ABSTRACT More than 180 susceptibility loci for breast cancer have been identified by genome-wide association studies (GWAS), mainly in Caucasian populations. However, many of these risk variants cannot be directly replicated in women of African ancestry, suggesting that causal variants are yet to be identified. Polygenic risk scores (PRS), which aggregate common genetic variants identified by GWAS, have been developed to predict genetic risk of breast cancer for Caucasian women, but there is no validated PRS for African American women. The linkage equilibrium in African ancestry populations is much less extensive than in Caucasian and Asian populations, which makes African ancestry population the ideal population to find causative variants after localizing a breast cancer susceptibility locus. Therefore, we propose a comprehensive analytical study that leverages several types of existing genetic datasets for breast cancer available to us and in public domains to address three specific aims. First, we aim to conduct cross-ethnic fine-mapping analysis for narrowing down casual variant candidate lists in 180+ loci of breast cancer. We will compile and harmonize genetic data from studies of breast cancer in women of African ancestry, including 7,525 cases and 6,207 controls, and leverage the association results from Caucasians (>122,000 cases and >105,000 controls), East Asians (>14,000 cases and >13,000 controls), and Latinos (4,400 cases and 7,500 controls). We will use a Bayesian statistical method to directly incorporate multiple functional annotations for the top variants in each locus. Second, we aim to develop breast cancer polygenic risk score models in African Americans by leveraging functional annotations, linkage disequilibrium, and gene expression data. Several PRSs will be developed for overall breast cancer risk and by estrogen receptor, cross-validated internally, and validated with external studies. Third, we aim to develop breast cancer risk prediction model by combining both genetic and non-genetic factors. The proposed study will efficiently utilize several types of existing data using innovative integrative approaches and has the potential to advance the field by narrowing down the genetic regions containing causal variants. More importantly, the risk prediction model has a good potential to translate knowledge from GWAS to the practice of breast cancer screening.

Abstract ? Administration Core A The Administration Core A will provide the necessary administrative and scientific oversight for the Program Project to ensure close interaction and effective integration of the Projects and Cores. More specifically, Core A will provide leadership to the primary components of the Program Project through an Executive Committee (EC), comprised of the principal investigators (PIs) of the Projects and Cores as well as senior investigators. The EC will oversee scientific progress, planning, and management and will be responsible for making decisions regarding scientific direction to promote and achieve scientific coherence. This Core will also be responsible for enhancing scientific interaction and communication amongst the PI?s and co-investigators of the Projects and Cores to ensure optimal utilization and integration of the clinical and genomic data across Projects. The Core will assemble an External Advisory Committee (EAC) comprised of scientists, clinicians and advocates representing the African American community to provide advice on scientific direction, recruitment strategies, as well as communication and dissemination of results with the broader community. Core A will also coordinate all program activities, including fiscal oversight of the individual Projects/Cores and subcontracts, the preparation of progress reports, regular communication with NCI, and the compliance with regulatory institutional and governmental policies regarding data sharing.

ABSTRACT An estimated 53% of U.S. adults have dyslipidemia, putting a majority of the U.S. adult population at high risk for related chronic diseases such as cardiovascular diseases, non-alcoholic fatty liver disease, and gallbladder disease. US Hispanic/Latinos (H/L) ages 18?74 have an overall prevalence of dyslipidemia of 65%, among the highest reported in the US. Lipid traits are highly heritable; estimates range from 20 to 70%, with common genetic variants explaining ~30% of the variance for these traits in Europeans. As serum concentrations of lipids are established therapeutic targets for many lipid-related chronic diseases, researchers have invested considerable effort into understanding the genetic epidemiology of lipid traits, however these large-scale efforts have almost exclusively considered Caucasians. Understudied at-risk populations provide a powerful design to gain insight into genetic mechanisms for disease because they can exhibit finer haplotypic structure and have different underlying causal variants. To ensure ancestrally diverse populations are not the last to benefit from the new era of precision medicine, we must both increase representation of ancestrally diverse populations in genetic research and develop expedited strategies for translating genomics for clinical utility. First, to enrich discovery, we will conduct the first large-scale GWAS and rare variant analyses for lipid- related traits in H/L. We will meta-analyze, fine-map, perform multivariate associations, and validate effects in all available H/L samples in analyses that will include >50,000 samples. Second, to interpret function, we will move GWAS findings into an interpretable biological context and characterize the regulatory mechanisms involved in lipid regulation via tissue-specific functional analysis, and ancestry-specific validation of effects using RNAseq data in two independent H/L cohorts. Identification of genes and pathways associated with lipid levels elucidates important basic biology about human metabolism, but isn?t necessarily clinically translatable. Thus, to evaluate clinical significance of lipid-associated genetic risk factors, we will use multiple massive genetic and electronic medical record repositories (including the Multiethnic Cohort, BioME, and BioVU) to identify clinical outcomes associated with single variants and genetically regulated expression of lipid- associated genes in H/L phenome-wide. Our design focuses effort on discovery of new variants and loci by pioneering genetic studies of lipid-related traits in diverse H/L populations, functional interpretation of variant effects via gene-based annotation and expression prediction with robust validation, and characterizing the clinical outcomes predicted by lipid-associated genetics in three large DNA bio-banks with linked electronic medical records. These population-specific, function- and outcome-oriented approaches will advance understanding of the genetic etiology of lipids and related traits with high H/L disparities of risk, revealing new biologic pathways and providing new avenues for precision treatment for H/L, a population that will constitute ~35% of the US population by the year 2050.

Several large-scale whole genome sequencing (WGS) and omics efforts are underway. Although these programs are making strides towards including ancestrally diverse populations, there is still a notable gap in analyses of these data that leverage diversity to empower discovery and improve our understanding of genotypic and phenotypic architecture across all populations. Substantial European bias persists in ongoing large scale WGS and omics efforts. Differences in genetic variation among ancestrally diverse groups are well-known, and although data are very limited, gene expression, methylation, and metabolites also differ among ancestrally diverse populations. The PAGE Study has been continuously funded by the NIH since 2008 to study genomic variation to advance our understanding of population architecture of complex genetic traits and diseases, particularly in the presence of ancestral diversity. We have established PAGE at the forefront of discovery and fine-mapping across >30 primary and >200 secondary complex traits and diseases in ancestrally diverse populations, and have led the way in array (e.g. the Multi-Ethnic Global Array, Global Screening Array) and statistical methods development specific to these diverse populations. Additionally, PAGE has served as a valuable resource to the scientific community, placing a high priority on quickly disseminating allele frequency data, GWAS summary statistics, study findings, and analytical software. PAGE Phase III (PAGE III) will include existing genetics and genomics data from more than 120,000 diverse individuals from six well characterized cohorts/biobanks. In PAGE III, we propose to extend and continue our invaluable work to date to 1) identify and characterize genetic variants that influence complex traits and diseases in ancestrally diverse individuals using both WGS data (n=42,000) and imputed genotyping data (n=124,000), 2) integrate information on sequence variation and omics to better understand the genetic underpinnings of complex traits in the diverse PAGE participants, and 3) characterize biological pathways underlying disease risk both within and between populations. For Aim 1, we will use newly generated sequence data to impute the largest sample of diverse participants every considered for the genetics of complex traits. For Aim 2, we will use newly existing data to impute gene expression, DNA methylation and metabolomics data into the remainder of PAGE samples, which will involve both extension and development of current methodology to ensure suitability for ancestrally diverse populations. Aim 3 will focus on integration of all omics data to inform discovery of novel pathways and the genetic basis of complex diseases and elucidate the molecular drivers of disease etiologies across diverse PAGE populations. As a continuation of our ground- breaking work in PAGE over the last decade, our proposal offers major advances towards understanding the genetic and genomic components underlying biological mechanisms of disease, and translational medicine in adversely affected and understudied diverse populations.

Abstract ? Project 2 Prostate cancer (PCa) incidence and mortality rates are highest in African American (AA) men compared to any other racial/ethnic population. These long-standing racial/ethnic differences have yet to be explained. Genome- wide association studies of PCa conducted in AA men have provided clear support for genetic differences in the allelic architecture of PCa across populations and strong support for a genetic basis underlying population differences in risk. There are also multiple lines of evidence to support a genetic contribution to aggressive/fatal PCa including recent sequencing studies that have provided support for rare (<1%) protein coding variation in contributing to aggressive PCa susceptibility. Cancer development and progression involves both the germline and somatic genomes however to date, discovery and characterization efforts of germline risk loci and somatic alterations in PCa have been conducted in isolation. We propose that large-scale efforts are warranted for the discovery of common and rare genetic variation that contributes to aggressive PCa susceptibility in AA men. In addition, through combining germline and somatic variation we expect to reveal novel biological mechanisms underlying PCa aggressiveness. In this Project, we seek to identify genetic factors that are associated with PCa aggressiveness in men of African ancestry. More specifically, in Aim 1, we plan to search for common germline variants associated with PCa risk profiles in a case-case GWAS of 5,700 cases with high-risk (stage T3/T4 or Gleason 8+ or PSA>20 ng/ml), 5,600 cases with intermediate-risk (stage T2b/T2c or Gleason 7 or PSA 10-20 ng/ml) and 3,800 cases with low-risk PCa (stage T1/T2a and Gleason ? 6 and PSA<10 ng/ml) from the RESPOND cohort and the African Ancestry PCa Consortium (AAPC). In Aim 2, we will conduct exome sequencing of these 15,000 PCa cases as well as 5,000 controls from AAPC to investigate the contribution of rare variation on PCa aggressiveness. In both Aims 1 and 2, single variant and gene burden association testing will be conducted by case-case comparisons of high- vs. low-risk disease, as well as by multi-nominal polytomous logistic regression by risk category. In Aim 3, we will test the relationship between germline genetics, lifestyle factors and somatic tumor characteristics (from Project 3) in AA men and assess whether germline genetics and lifestyle factors interact with somatic profiles and transcription states to distinguish PCa risk profiles. We expect these integrated germline-risk factor-somatic analyses to reveal tumor subgroups and shared biological mechanisms that are clinically important in the future prediction of disease progression and survival. We expect the findings from this Project to significantly advance knowledge of susceptibility to aggressive PCa and racial/ethnic disparities in PCa risk, and to guide the development of future preventive, early detection and prognostic measures for AA men.

Abstract ? Program Project Overview African American (AA) men have a >60% higher incidence and are more likely to be diagnosed with aggressive PCa than white men. Reasons for the greater burden of aggressive disease in AA men are unknown, but are likely to include a multitude of factors including social factors such as lifetime stress, inherited susceptibility, and tumor-related features such as somatic alterations and local inflammation in the microenvironment. The overarching goal of this Program Project is to uncover the social and biological factors that are related to PCa aggressiveness in AA men. To accomplish this objective, we will establish a large, national, population-based cohort study, RESPOND, (Research on Prostate Cancer in Men of African Ancestry: Defining the Roles of Genetics, Immunity and Social Stress) of 10,000 AA men with incident PCa identified through nine SEER and NPCR U.S. cancer registries from states that include 38% of all AA PCa cases in the U.S.. The cohort will provide comprehensive information on multilevel stressors over the lifecourse such as discrimination, early-life adversity, and neighborhood disorder, including geospatial neighborhood data over time and degree of perceived stress; 2) lifestyle factors and health behaviors; 3) disease-specific factors including PSA screening history and treatment choice; 4) germline DNA to study genetic susceptibility, and 5) tumor block samples for characterization of somatic variation and immune profiling of the tumor microenvironment. No previous study has attempted to obtain information across these domains in a single large sample in order to understand the relative contribution of each and relationships between molecular and non-genetic components. In order to address these goals, we have assembled a multi-disciplinary team of scientists and clinicians with established track records in PCa research. Leveraging the RESPOND resource and investigator expertise, we have designed a Program Project composed of four Projects that are supported by four Cores which are all focused on the central theme of identifying social and biological factors related to PCa disease aggressiveness in AA men. These Projects include: the investigation of multilevel social stressors across the lifecourse in relationship with aggressive PCa (Project 1); genome-wide discovery efforts of germline susceptibility loci for aggressive PCa and examination of the relationship between germline and somatic variation (Project 2); the identification of underlying somatic alterations in PCa tumors and biological pathways that are related to aggressive disease (Project 3); and, a detailed assessment of inflammation in the tumor microenvironment as it relates to PCa aggressiveness in AA men (Project 4). Each of the four Projects address a distinct research domain, however, when studied together, create scientific synergy and a far more comprehensive picture of the major factors that contribute to aggressive PCa in AA men. The information we will discover is likely to have immediate clinical implications in the areas of improved patient stratification and personalized medicine. Hence, this study has broad reaching significance and addresses numerous challenges in the clinical management of PCa in AA men.

At the root of health disparities is the diversity in the biological, social, cultural, behavioral, and economic characteristics of individuals and populations. The study of ethnic/racial disparities begins with an understanding of the characteristics of each group and how these characteristics affect cancer development and outcomes. This new application for a joint T32 postdoctoral training program at University of Hawaii Cancer Center (UHCC) and the University of Southern California (USC) Norris Comprehensive Cancer Center (NCCC) addresses cancer health disparities utilizing a unique resource, the Multiethnic Cohort (MEC) Study, a population-based prospective cohort study that has followed African Americans, Native Hawaiians, Japanese Americans, Latinos, and whites for the past 25 years. Linkages to cancer registries, Medicare, and other data sources provide a variety of endpoints for MEC and many other cancer disparity studies conducted at both institutions. The complementary research at UHCC and USC is supported by cutting edge cores and benefits from many years of experience. The specific aims of the proposed training program are: 1. To train postdoctoral fellows for independent careers in cancer research in diverse populations, with a specific focus on the sociocultural, nutritional, lifestyle and biological risk factors leading to cancer health disparities in a variety of ethnic/racial minorities. This program is highly unique and will leverage the long- standing (30+ years) collaboration between research programs at two NCI-designated Cancer Centers to permit a broader experience in training and tools for epidemiological research. 2. To provide extensive applied research experience in a multidisciplinary environment focusing on nutritional, molecular, genetic and translational epidemiology that builds on multiple unique racial and ethnically diverse studies of cancer. Together with strong mentorship, these studies will serve as the centerpiece for an effective transdisciplinary training, which will allow trainees to design innovative high-impact projects, and will prepare trainees for careers in cancer health disparities research. During the 5-year grant period, the training program will educate 8 postdoctoral fellows in Hawaii and 10 in Los Angeles. All trainees will be exposed to a wide range of innovative molecular and traditional cancer prevention approaches (epidemiology, biomarkers, genomics, and nutrition) to address a wide range of cancer disparity research questions across the ethnic/racial populations in Hawaii and California through the close cooperation of mentors in both locations and regular exchange visits. The dual-site training, which regular includes exchange visits, is a particular strength of the proposed training program as it will provide exposure to multi- institutional and multi-disciplinary collaborative research that builds on the expertise and long-standing collaboration between UH and USC in research, education and training.

PROJECT SUMMARY / ABSTRACT Native Hawaiian women (NatHW) experience substantial health disparities compared to women of other races/ethnicities, evidenced by their short healthy life expectancy and high all-cause and cancer mortality rates. Reasons for NatHW?s health disparities likely involve a combination of causes, including genetic susceptibility, individual and contextual socioeconomic status (SES), obesity and related lifestyle factors, smoking, healthcare access and quality, and the environment. The multifactorial basis of disparities requires an array of information and integrative approaches in research. Herein, we propose to use epigenetic biomarkers, known to reflect and mediate the combined effects of some of the genetic and environmental stressors, to enhance our mechanistic understanding of NatHW?s disparities in key carcinogenic metabolic traits. We will base the study in the Multiethnic Cohort Study (MEC), which is following the largest number of otherwise understudied Native Hawaiians (~7,850 women, ~6,120 men), as well as >201,000 older adults of mainly four other ethnicities (African American, Japanese American, Latino, white) since 1993-1996 with comprehensive characterization. Research to date indicates that there are significant sex and racial/ethnic differences in the human DNA methylome and in DNA methylation age (DNAm-age) acceleration. Thus, to enable an initial study in the MEC of leukocyte DNA methylation as a likely mediator of NatHW?s health disparities, we request a supplement to the MEC grant (U01 CA164973) in response to NOT-OD-20-048. We will analyze the archived leukocyte DNA samples in the Adiposity Phenotype Study (APS), which is a cross-sectional subset of the MEC, using the Illumina HumanMethylation EPIC bead chips. This will allow us to compare genome-wide methylation patterns and DNAm-age parameters in 140 NatHW and 140 white women, aged 60-77 years at the time of blood collection. We plan to (Aim 1) compare the DNA methylation patterns by constructing a DNA methylation score (DNAm- score) for NatHW vs. white women using penalized logistic regression of ethnicity on methylation at CpGs, adjusted for % Native Hawaiian genetic ancestry. Based on the selected differentially methylated genes, we will infer the biological functions in pathway analysis. We hypothesize that we will discover a DNA methylation signature and accelerated DNAm-age for NatHW vs. white women. We will then (Aim 2) quantify the proportion of the excess metabolic risk in NatHW compared to white women, namely higher insulin, triglycerides, visceral fat, and liver fat, respectively, mediated through the DNA methylome, based on the DNAm-score and DNAm- age. Finally, we will (Aim 3) identify behavioral and environmental determinants of the DNAm-score or DNAm- age among the wealth of data in the MEC-APS (adulthood weight gain, total body fat, lifetime smoking, diet quality, alcohol and meat intake, physical in/activity, sleep hours, and individual and neighborhood SES). This study addresses the goal of the U3 supplement and the MEC U01 grant and may yield important insight into the mechanisms of the disparities in NatHW.

Abstract Prostate cancer (PCa) incidence is highest in African Americans and lowest in Asians. These long-standing racial/ethnic differences have yet to be explained. Genome-wide association studies of PCa have provided support for common and population-specific genetic effects for PCa and for a genetic basis of the underlying population differences in risk. To further progress in understanding the genetic basis of PCa across populations, we propose to substantially augment the size of genetic association studies in men of European, African, Asian and Latino ancestry to create the largest genetic database of PCa ever assembled in these populations, with substantially greater statistical power for novel discovery of risk alleles for PCa as well as aggressive disease. More specifically, we will expand studies in men of African ancestry from 10,368 cases and 10,986 controls to 34,000 cases and 74,000 controls, in men of Asian ancestry from 8,610 cases and 18,809 controls to 20,000 cases and 40,000 controls, in men of Latino ancestry from 2,714 cases and 5,239 controls to 10,000 cases and 20,000 controls, and in men of European ancestry from 82,000 cases and 61,000 controls to 117,000 cases and 517,000 controls, with all studies imputed to a multiethnic whole-genome sequence reference panel (e.g. TOPMed). In Aim 1, we will search for novel common risk alleles for overall and aggressive PCa in ethnic-specific and multiethnic analyses. Within known and newly discovered risk regions, we will conduct multiethnic fine- mapping using novel Bayesian statistical approaches that incorporate functional annotations and biology with statistical evidence to identify independent markers of risk as well as the most promising functional candidates. In Aim 2, we will create the first multiethnic genome-wide SNP-eQTL prostate reference panel for men of European, African, Asian and Latino ancestry using whole-transcriptome RNA sequencing of ~1,000 histologically normal, fresh-frozen prostate tissue specimens. We will characterize eQTLs in the sample and impute gene expression in men of European, African, Asian, and Latino ancestry to perform a multiethnic transcriptome-wide association scan (TWAS). In Aim 3, we will construct and evaluate a polygenic risk score (PRS) across populations, using known and novel risk variants from Aim 1 and TWAS loci from Aim 2. PRS validation testing will be conducted in three independent multiethnic cohorts (>38,000 PCa cases from ATLAS, All of Us and CCPM). We expect findings from this study will make a major contribution to our understanding of genetic susceptibility to PCa and lead to better risk models that more accurately predict a man's risk of developing PCa and are efficacious across racial/ethnic populations.