Gene-level Association Research Articles

Composite Kernel Association Test (CKAT) for SNP-set joint assessment of genotype and genotype-by-treatment interaction in Pharmacogenetics studies.

It is of substantial interest to discover novel genetic markers that influence drug response in order to develop personalized treatment strategies that maximize therapeutic efficacy and safety. To help enable such discoveries, we focus on testing the association between the cumulative effect of multiple single nucleotide polymorphisms (SNPs) in a particular genomic region and a drug response of interest. However, the currently existing methods are either computational inefficient or not able to control type I error and provide decent power for whole exome or genome analysis in Pharmacogenetics (PGx) studies with small sample sizes. In this article, we propose the Composite Kernel Association Test (CKAT), a flexible and robust kernel machine-based approach to jointly test the genetic main effect and SNP-treatment interaction effect for SNP-sets in Pharmacogenetics (PGx) assessments embedded within randomized clinical trials. An analytic procedure is developed to accurately calculate the P-value so that computationally extensive procedures (e.g. permutation or perturbation) can be avoided. We evaluate CKAT through extensive simulation studies and application to the gene-level association test of the reduction in Clostridium difficile infection recurrence in patients treated with bezlotoxumab. The results demonstrate that the proposed CKAT controls type I error well for PGx studies, is efficient for whole exome/genome association analysis and provides better power performance than existing methods across multiple scenarios. The R package CKAT is publicly available on CRAN https://cran.r-project.org/web/packages/CKAT/. Supplementary data are available at Bioinformatics online.

Integrative Omics Approach to Identifying Genes Associated With Atrial Fibrillation.

Rationale: GWAS (Genome-Wide Association Studies) have identified hundreds of genetic loci associated with atrial fibrillation (AF). However, these loci explain only a small proportion of AF heritability. Objective: To develop an approach to identify additional AF-related genes by integrating multiple omics data. Methods and Results: Three types of omics data were integrated: (1) summary statistics from the AFGen 2017 GWAS; (2) a whole blood EWAS (Epigenome-Wide Association Study) of AF; and (3) a whole blood TWAS (Transcriptome-Wide Association Study) of AF. The variant-level GWAS results were collapsed into gene-level associations using fast set-based association analysis. The CpG-level EWAS results were also collapsed into gene-level associations by an adapted SNP-set Kernel Association Test approach. Both GWAS and EWAS gene-based associations were then meta-analyzed with TWAS using a fixed-effects model weighted by the sample size of each data set. A tissue-specific network was subsequently constructed using the NetWAS (Network-Wide Association Study). The identified genes were then compared with the AFGen 2018 GWAS that contained more than triple the number of AF cases compared with AFGen 2017 GWAS. We observed that the multiomics approach identified many more relevant AF-related genes than using AFGen 2018 GWAS alone (1931 versus 206 genes). Many of these genes are involved in the development and regulation of heart- and muscle-related biological processes. Moreover, the gene set identified by multiomics approach explained much more AF variance than those identified by GWAS alone (10.4% versus 3.5%). Conclusions: We developed a strategy to integrate multiple omics data to identify AF-related genes. Our integrative approach may be useful to improve the power of traditional GWAS, which might be particularly useful for rare traits and diseases with limited sample size.

Gene-level genome-wide association analysis of suicide attempt, a preliminary study in a psychiatric Mexican population.

BackgroundEvidence suggests that liability for suicide behavior is heritable; additionally, suicide has been partly related to other psychiatric disorders. Nevertheless, most of the information reported so far address Caucasian and Asian individuals. Hence, our aim was to conduct a gene‐level association study in Mexican psychiatric individuals diagnosed with suicide attempt.MethodsWe recruited 192 individuals from two clinical centers in Mexico. All participants were born in Mexico and had Mexican parents and grandparents. Direct genotyping was performed using the commercial platform Infinium PsychArray BeadChip. A p‐value lower than 1e‐05 was considered as gene‐level significant and a p‐value lower than 1e‐04 was considered as gene‐level nominal significant.ResultsOur analyses showed that SCARA5 was associated to suicide intent at a gene‐level with statistical significance (p‐value = 1.12e‐6). Other genes were nominally associated with suicide attempt: GHSR (p‐value = 0.0004), RGS10 (p‐value = 5.13e‐5), and STK33 (p‐value = 3.62e‐5). Regarding gene variant analyses, the SNPs with a statistical association (p > .05) were rs561361616, rs1537577, rs11198999 for RGS10, and rs11041981, rs11041993, rs11041994, rs11041995, rs11041997, rs10840083, rs10769918 for STK33. For these genes, previous studies have associated SCARA5 with depression, GHSR with alcohol dependence and depression, and RGS10 with schizophrenia and depression. To date, STK33 has not been associated with any psychiatric disorder.ConclusionOur outcomes revealed that SCARA5, GHSR, RGS10 and STK33 could be considered as risk biomarkers for suicide attempt behavior in our Mexican psychiatric sample. We recommend to perform larger scale analyses to have conclusive results.

Advancements of transcriptome imputation and related transcriptome-wide association studies

Despite the progresses of genome-wide association studies (GWASs) in revealing genetic mechanisms of human complex traits, the basis through which most identified risk variants function are highly unknown and need further investigations as well as discoveries. Recent advancements of transcriptome predictions put the transcriptome-wide association studies (TWASs) forward into a new era. TWAS through imputed transcriptomes could discover more gene-trait associations and relevant joint-tissue TWAS via eQTL analysis provide insights into furthering elucidations about gene-level association studies in difficult-to-acquire tissues. This mini-review goes over the recent advancements of gene expression imputations as well as the gene-trait association studies, which highlight the importance of genetically regulated expression (GReX) in this area.

Identification of transcriptional isoforms associated with survival in cancer patient

The Cancer Genome Atlas (TCGA) project produced RNA-Seq data for tens of thousands of cancer and non-cancer samples with clinical survival information, providing an unprecedented opportunity for analyzing prognostic genes and their isoforms. In this study, we performed the first large-scale identification of transcriptional isoforms that are specifically associated with patient prognosis, even without gene-level association. These specific isoforms are defined as Transcripts Associated with Patient Prognosis (TAPPs). Although a group of TAPPs are the principal isoforms of their genes with intact functional protein domains, another group of TAPPs lack important protein domains found in their canonical gene isoforms. This dichotomy in the distribution of protein domains may indicate different patterns of TAPPs association with cancer. TAPPs in protein-coding genes, especially those with altered protein domains, are rich in known cancer driver genes. We further identified multiple types of cancer recurrent TAPPs, such as DCAF17-201, providing a new approach for the detection of cancer-associated events. In order to make the wide research community to study prognostic isoforms, we developed a portal named GESUR (http://gesur.cancer-pku.cn/), which illustrates the detailed prognostic characteristics of TAPPs and other isoforms. Overall, our integrated analysis of gene expression and clinical parameters provides a new perspective for understanding the applications of different gene isoforms in tumor progression.

A meta-analysis of genome-wide association studies identifies multiple longevity genes

Human longevity is heritable, but genome-wide association (GWA) studies have had limited success. Here, we perform two meta-analyses of GWA studies of a rigorous longevity phenotype definition including 11,262/3484 cases surviving at or beyond the age corresponding to the 90th/99th survival percentile, respectively, and 25,483 controls whose age at death or at last contact was at or below the age corresponding to the 60th survival percentile. Consistent with previous reports, rs429358 (apolipoprotein E (ApoE) ε4) is associated with lower odds of surviving to the 90th and 99th percentile age, while rs7412 (ApoE ε2) shows the opposite. Moreover, rs7676745, located near GPR78, associates with lower odds of surviving to the 90th percentile age. Gene-level association analysis reveals a role for tissue-specific expression of multiple genes in longevity. Finally, genetic correlation of the longevity GWA results with that of several disease-related phenotypes points to a shared genetic architecture between health and longevity.

A gene based approach to test genetic association based on an optimally weighted combination of multiple traits.

There is increasing evidence showing that pleiotropy is a widespread phenomenon in complex diseases for which multiple correlated traits are often measured. Joint analysis of multiple traits could increase statistical power by aggregating multiple weak effects. Existing methods for multiple trait association tests usually study each of the multiple traits separately and then combine the univariate test statistics or combine p-values of the univariate tests for identifying disease associated genetic variants. However, ignoring correlation between phenotypes may cause power loss. Additionally, the genetic variants in one gene (including common and rare variants) are often viewed as a whole that affects the underlying disease since the basic functional unit of inheritance is a gene rather than a genetic variant. Thus, results from gene level association tests can be more readily integrated with downstream functional and pathogenic investigation, whereas many existing methods for multiple trait association tests only focus on testing a single common variant rather than a gene. In this article, we propose a statistical method by Testing an Optimally Weighted Combination of Multiple traits (TOW-CM) to test the association between multiple traits and multiple variants in a genomic region (a gene or pathway). We investigate the performance of the proposed method through extensive simulation studies. Our simulation studies show that the proposed method has correct type I error rates and is either the most powerful test or comparable with the most powerful tests. Additionally, we illustrate the usefulness of TOW-CM based on a COPDGene study.

Powerful statistical method to detect disease‐associated genes using publicly available genome‐wide association studies summary data

Genome-wide association studies (GWAS) have thus far achieved substantial success. In the last decade, a large number of common variants underlying complex diseases have been identified through GWAS. In most existing GWAS, the identified common variants are obtained by single marker-based tests, that is, testing one single-nucleotide polymorphism (SNP) at a time. Generally, the basic functional unit of inheritance is a gene, rather than a SNP. Thus, results from gene-level association test can be more readily integrated with downstream functional and pathogenic investigation. In this paper, we propose a general gene-based p-value adaptive combination approach (GPA) which can integrate association evidence of multiple genetic variants using only GWAS summary statistics (either p-value or other test statistics). The proposed method could be used to test genetic association for both continuous and binary traits through not only one study but also multiple studies, which would be helpful to overcome the limitation of existing methods that can only be applied to a specific type of data. We conducted thorough simulation studies to verify that the proposed method controls type I errors well, and performs favorably compared to single-marker analysis and other existing methods. We demonstrated the utility of our proposed method through analysis of GWAS meta-analysis results for fasting glucose and lipids from the international MAGIC consortium and Global Lipids Consortium, respectively. The proposed method identified some novel trait associated genes which can improve our understanding of the mechanisms involved in -cell function, glucose homeostasis, and lipids traits.

Recent advances and perspectives in next generation sequencing application to the genetic research of type 2 diabetes.

Type 2 diabetes (T2D) mellitus is a common complex disease that currently affects more than 400 million people worldwide and has become a global health problem. High-throughput sequencing technologies such as whole-genome and whole-exome sequencing approaches have provided numerous new insights into the molecular bases of T2D. Recent advances in the application of sequencing technologies to T2D research include, but are not limited to: (1) Fine mapping of causal rare and common genetic variants; (2) Identification of confident gene-level associations; (3) Identification of novel candidate genes by specific scoring approaches; (4) Interrogation of disease-relevant genes and pathways by transcriptional profiling and epigenome mapping techniques; and (5) Investigation of microbial community alterations in patients with T2D. In this work we review these advances in application of next-generation sequencing methods for elucidation of T2D pathogenesis, as well as progress and challenges in implementation of this new knowledge about T2D genetics in diagnosis, prevention, and treatment of the disease.

Abstract 1588: Germline variation in DNA repair genes and risk of Barrett's esophagus and esophageal adenocarcinoma

Abstract Incidence of esophageal adenocarcinoma (EA) has risen substantially in Western countries over recent decades. Established risk factors for EA and its precursor lesion, Barrett’s esophagus (BE), include reflux, obesity, and tobacco smoking. Inherited genetic variation also influences disease risk, although only a limited number of susceptibility loci have been identified. Genomic analyses of EA tumors have revealed a distinctive mutational signature, high mutational burden, and extensive somatic chromosome alterations, features also observed in high-risk BE tissue. To explore whether germline variation in DNA repair-related genes may be associated with altered disease susceptibility, we analyzed data from a recent meta-analysis of genome-wide association studies (GWAS) encompassing 4,112 EA cases, 6,167 BE cases, and 17,159 controls, representing the largest sample size assembled for these conditions. Using a gene-based testing approach (VEGAS2), we assessed 263 DNA repair-related genes and found that variation in NEIL2, a mediator of base excision repair (BER), was significantly associated with risk of BE (P=1.4×10-5, q<0.05). No other gene-level associations with BE or EA survived correction for multiple comparisons. SNP-level analysis of 239 polymorphisms at the NEIL2 locus revealed six variants strongly associated with altered risk of BE (P<5×10-5, q<0.01), with the index SNP classified as an intronic eQTL for NEIL2 in esophageal tissue. Four of these SNPs were also associated with risk of EA (P<0.05), with odds ratios in the same direction and of similar magnitude. Our results provide evidence that germline genetic variation in a DNA glycosylase enzyme (NEIL2) may influence risk of BE/EA, and suggest a potential novel biological role for altered BER in BE/EA pathogenesis. Citation Format: Matthew F. Buas, Li Yan, Xuan Peng, Qianya Qi, Jianhong Chen, Aaron Thrift, Qianchuan He, Lynn Onstad, Puya Gharahkhani, Stuart MacGregor, Thomas L. Vaughan, Margaret M. Madeleine. Germline variation in DNA repair genes and risk of Barrett's esophagus and esophageal adenocarcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 1588.

Abstract 4254: Integrated analysis with MVisAGe identifies concordant and discordant genomic alterations of driver genes

Abstract Integrated analyses of multiple genomic datatypes are now common in cancer profiling studies. Such data present opportunities for numerous computational experiments, yet analytic pipelines are limited. Tools such as the cBioPortal and Regulome Explorer, while useful, are not easy to access programmatically or implement locally. The MVisAGe R package allows users to quantify gene-level associations between two genomic datatypes in order to investigate the effect of genomic alterations (e.g. DNA copy number changes on gene expression). Visualizing Pearson/Spearman correlation coefficients according to the genomic positions of the underlying genes provides a powerful yet novel tool for conducting exploratory analyses. We demonstrate its utility by analyzing three publicly available datasets from squamous tumors. Our approach highlights canonical oncogenes in chr11q13 that displayed the strongest associations between expression and copy number, including CCND1 and CTTN, genes not identified by copy number analysis in the primary reports. We demonstrate highly concordant usage of shared oncogenes on chr3q, yet strikingly diverse oncogene usage on chr11q as a function of HPV infection status. Regions of chr19 that display remarkable associations between methylation and gene expression were identified, as were previously unreported miRNA-gene expression associations that may contribute to the epithelial-to-mesenchymal transition. Citation Format: Vonn Walter, Ying Du, Ludmila Danilova, Michele Hayward, D. Neil Hayes. Integrated analysis with MVisAGe identifies concordant and discordant genomic alterations of driver genes [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 4254.

Genome-wide association study of multisite chronic pain in UK Biobank.

Chronic pain is highly prevalent worldwide and represents a significant socioeconomic and public health burden. Several aspects of chronic pain, for example back pain and a severity-related phenotype ‘chronic pain grade’, have been shown previously to be complex heritable traits with a polygenic component. Additional pain-related phenotypes capturing aspects of an individual’s overall sensitivity to experiencing and reporting chronic pain have also been suggested as a focus for investigation. We made use of a measure of the number of sites of chronic pain in individuals within the UK general population. This measure, termed Multisite Chronic Pain (MCP), is a complex trait and its genetic architecture has not previously been investigated. To address this, we carried out a large-scale genome-wide association study (GWAS) of MCP in ~380,000 UK Biobank participants. Our findings were consistent with MCP having a significant polygenic component, with a Single Nucleotide Polymorphism (SNP) heritability of 10.2%. In total 76 independent lead SNPs at 39 risk loci were associated with MCP. Additional gene-level association analyses identified neurogenesis, synaptic plasticity, nervous system development, cell-cycle progression and apoptosis genes as enriched for genetic association with MCP. Genetic correlations were observed between MCP and a range of psychiatric, autoimmune and anthropometric traits, including major depressive disorder (MDD), asthma and Body Mass Index (BMI). Furthermore, in Mendelian randomisation (MR) analyses a causal effect of MCP on MDD was observed. Additionally, a polygenic risk score (PRS) for MCP was found to significantly predict chronic widespread pain (pain all over the body), indicating the existence of genetic variants contributing to both of these pain phenotypes. Overall, our findings support the proposition that chronic pain involves a strong nervous system component with implications for our understanding of the physiology of chronic pain. These discoveries may also inform the future development of novel treatment approaches.

The association of germline variants with chronic lymphocytic leukemia outcome suggests the implication of novel genes and pathways in clinical evolution

BackgroundChronic Lymphocytic Leukemia (CLL) is the most frequent lymphoproliferative disorder in western countries and is characterized by a remarkable clinical heterogeneity. During the last decade, multiple genomic studies have identified a myriad of somatic events driving CLL proliferation and aggressivity. Nevertheless, and despite the mounting evidence of inherited risk for CLL development, the existence of germline variants associated with clinical outcomes has not been addressed in depth.MethodsExome sequencing data from control leukocytes of CLL patients involved in the International Cancer Genome Consortium (ICGC) was used for genotyping. Cox regression was used to detect variants associated with clinical outcomes. Gene and pathways level associations were also calculated.ResultsSingle nucleotide polymorphisms in PPP4R2 and MAP3K4 were associated with earlier treatment need. A gene-level analysis evidenced a significant association of RIPK3 with both treatment need and survival. Furthermore, germline variability in pathways such as apoptosis, cell-cycle, pentose phosphate, GNα13 and Nitric oxide was associated with overall survival.ConclusionOur results support the existence of inherited conditionants of CLL evolution and points towards genes and pathways that may results useful as biomarkers of disease outcome. More research is needed to validate these findings.

Exome sequencing of 20,791\xa0cases of type 2 diabetes and 24,440\xa0controls

Protein-coding genetic variants that strongly affect disease risk can yield relevant clues to disease pathogenesis. Here we report exome-sequencing analyses of 20,791 individuals with type 2 diabetes (T2D) and 24,440 non-diabetic control participants from 5 ancestries. We identify gene-level associations of rare variants (with minor allele frequencies of less than 0.5%) in 4 genes at exome-wide significance, including a series of more than 30 SLC30A8 alleles that conveys protection against T2D, and in 12 gene sets, including those corresponding to T2D drug targets (P = 6.1 × 10−3) and candidate genes from knockout mice (P = 5.2 × 10−3). Within our study, the strongest T2D gene-level signals for rare variants explain at most 25% of the heritability of the strongest common single-variant signals, and the gene-level effect sizes of the rare variants that we observed in established T2D drug targets will require 75,000–185,000 sequenced cases to achieve exome-wide significance. We propose a method to interpret these modest rare-variant associations and to incorporate these associations into future target or gene prioritization efforts.

Candidate SNPs enhance prediction of cognitive impairment after blood or marrow transplantation (BMT) for hematologic malignancy (HM).

11520 Background: We tested the hypothesis that candidate genetic variants are associated with cognitive impairment in BMT recipients for HM, and that inclusion of genetic variants improves the performance of a risk prediction model that includes only clinical and demographic characteristics. Methods: We used standardized tests to assess cognitive function in 277 adult BMT recipients at City of Hope (COH). Global Deficit Score (GDS) ≥0.50 was used as indicator of cognitive impairment. Generalized estimating equation models and logic regression were used to identify single-SNP and gene-level associations with cognitive impairment post-BMT. Three risk prediction models were developed in the COH cohort using elastic net regression: Base Model (sociodemographics); Clinical Model (Base Model + clinical characteristics, therapeutic exposures and baseline cognitive reserve); Combined Model (Clinical + Genetic Model). The Genetic Model included significant SNPs in blood brain barrier, telomere homeostasis and DNA repair identified from single- and gene-level analyses. Models were validated in an independent cohort of long-term BMT survivors (BMTSS) with (n = 141) and without (n = 258) memory problems. Results: Training set (COH): The cohort included 58.5% males; 68.6% non-Hispanic whites; 46.6% allogeneic BMT recipients; median age at BMT: 51.6y. The mean area under the receiver operating characteristic curve (AUC) was: Base Model: 0.69 (95%CI: 0.63-0.75); Clinical Model: 0.77 (95%CI: 0.71-0.83); Combined Model: 0.89 (95%CI: 0.84-0.92). Test set (BMTSS): Median age at BMT was 45y; 53.5% were males; 88.4% non-Hispanic whites. Testing the models in BMTSS yielded mean AUC of 0.57 (95%CI: 0.49-0.63) in the Clinical Model and 0.72, (95%CI: 0.65-0.78) in the Combined Model. Conclusions: These findings provide evidence on the utility of a validated risk prediction tool that incorporates genetic factors that could identify BMT recipients at risk for cognitive impairment, providing opportunities for targeted interventions.

OR05-1 Genetic Discovery and Translational Decision Support from Exome Sequencing of 45,231 Type 2 Diabetes Cases and Controls from Five Ancestries

Analysis of human exome sequences, the protein-coding regions of genomes, has the potential to identify variation in genes relevant to disease for expedited clinical and therapeutic translation. Here we present results from a massive-scale type 2 diabetes (T2D) exome sequencing project. We analyzed exomes from 20,791 participants with T2D and 24,440 controls from five ancestries (Hispanic/Latino 33.8%, European 24.6%, African-American 13.9%, East Asian 14.1%, South Asian 13.6%) with mean depth 40x, and conducted both single-variant and gene-level association testing. Eighteen single-variant signals reached exome-wide significance (p< 4.3×10-7); all were common variants (minor allele frequency [MAF] > 5%) except MC4R p.Ile269Asn, which was almost exclusively seen in Hispanic/Latinos with MAF 0.89% and T2D OR=2.17 (95% CI: 1.63-2.89) in that population. Only one single-variant association represented a previously unknown T2D genetic locus (SFI1 p.Arg724Trp), but this finding failed to replicate in an independent cohort. We identified gene-level associations composed of rare (MAF < 0.5%) variants (a) in three genes each reaching a gene-level exome-wide significance threshold of 6.57×10-7 (MC4R, PAM, and SLC30A8), including a T2D protective series of >30 SLC30A8 alleles, and (b) within 12 gene sets, including those corresponding to known T2D drug targets (p=6.1×10-3) and candidate genes from knockout mice (p=5.2×10-3). These aggregate gene-level associations suggest that larger sample sizes will allow more refined identification of novel T2D genes, including potential new drug targets. Based on the rare-variant gene-level effect sizes we observed in established T2D drug targets, we estimate that 110K-180K sequenced cases would be required for reliable identification of new T2D drug targets. Additionally, we have developed a Bayesian framework using our association results to facilitate prioritization of genes for translational decision support.

MON-186 An Evaluation of the Ability of Current Exome Sequence Datasets to Retrospectively Validate Drugs for T2D or Related Metabolic Traits

Human genetics offers the potential to inform on the viability of therapeutic targets through the identification of associations between naturally occurring genetic variants and clinical outcomes. Whole Exome Sequencing (WES) enables comprehensive analysis of protein coding variants, which can directly link genes to disease or related traits. We hypothesized that large-scale WES analysis could retrospectively (a) confirm the efficacy and (b) predict the side effects of known drugs for type 2 diabetes (T2D) or related metabolic traits. We analyzed WES data for 20,341 T2D cases and 23,981 controls with additional lipid, body mass index (BMI), and blood pressure (BP) measurements. Eight sets of curated genes were either (a) targeted by a medication for T2D (8 genes), hypercholesterolemia (13 genes), obesity (6 genes), or high BP (41 genes); or (b) targeted by a medication with a known side effect of T2D (8 genes), hypercholesterolemia (54 genes), obesity (230 genes), or elevated BP (126 genes). Rare variant gene-level burden tests were conducted for each gene and trait, and then a Mann-Whitney U test was used to assess whether each set of genes exhibited in aggregate stronger-than-expected gene-level associations for its corresponding trait. We observed several significant set-level associations for gene sets defined on the basis of drug primary effects, including for known T2D drug targets (T2D p = 0.002300) and for known cholesterol drug targets (total cholesterol p = 0.003790; low-density lipoprotein p = 0.00835). However, significant gene-level associations were not observed between known obesity drug targets and BMI (p = 0.3418) or between known BP targets and BP (p = 0.1597). When analyzing gene sets defined on the basis of drug side effects, we observed a significant set-level hypercholesterolemia association for targets of drugs with a side effect of hypercholesterolemia (p = 0.01614), but did not observe a significant set-level BMI association for genes targeted by drugs with a side effect of obesity (p = 0.6577) or a significant set-level BP association for genes targeted by drugs with a side effect of elevated BP (p = 0.3551). This analysis demonstrates the feasibility of current exome sequence sample sizes to predict primary and secondary side effects of drugs related to T2D or lipid traits, and suggests that drugs related to BMI or BP may be more challenging to validate with existing WES data. Drugs initially approved for a non-cholesterol and non-diabetes indication, but whose target demonstrates a significant association with lipids or T2D in our data, may be potential candidates for drug repurposing.

Genetic variant predictors of gene expression provide new insight into risk of colorectal cancer

Genome-wide association studies have reported 56 independently associated colorectal cancer (CRC) risk variants, most of which are non-coding and believed to exert their effects by modulating gene expression. The computational method PrediXcan uses cis-regulatory variant predictors to impute expression and perform gene-level association tests in GWAS without directly measured transcriptomes. In this study, we used reference datasets from colon (n = 169) and whole blood (n = 922) transcriptomes to test CRC association with genetically determined expression levels in a genome-wide analysis of 12,186 cases and 14,718 controls. Three novel associations were discovered from colon transverse models at FDR ≤ 0.2 and further evaluated in an independent replication including 32,825 cases and 39,933 controls. After adjusting for multiple comparisons, we found statistically significant associations using colon transcriptome models with TRIM4 (discovery P = 2.2 × 10− 4, replication P = 0.01), and PYGL (discovery P = 2.3 × 10− 4, replication P = 6.7 × 10− 4). Interestingly, both genes encode proteins that influence redox homeostasis and are related to cellular metabolic reprogramming in tumors, implicating a novel CRC pathway linked to cell growth and proliferation. Defining CRC risk regions as one megabase up- and downstream of one of the 56 independent risk variants, we defined 44 non-overlapping CRC-risk regions. Among these risk regions, we identified genes associated with CRC (P < 0.05) in 34/44 CRC-risk regions. Importantly, CRC association was found for two genes in the previously reported 2q25 locus, CXCR1 and CXCR2, which are potential cancer therapeutic targets. These findings provide strong candidate genes to prioritize for subsequent laboratory follow-up of GWAS loci. This study is the first to implement PrediXcan in a large colorectal cancer study and findings highlight the utility of integrating transcriptome data in GWAS for discovery of, and biological insight into, risk loci.

Improved Pathogenic Variant Localization via a Hierarchical Model of Sub-regional Intolerance

Different parts of a gene can be of differential importance to development and health. This regional heterogeneity is also apparent in the distribution of disease-associated mutations, which often cluster in particular regions of disease-associated genes. The ability to precisely estimate functionally important sub-regions of genes will be key in correctly deciphering relationships between genetic variation and disease. Previous methods have had some success using standing human variation to characterize this variability in importance by measuring sub-regional intolerance, i.e., the depletion in functional variation from expectation within a given region of a gene. However, the ability to precisely estimate local intolerance was restricted by the fact that only information within a given sub-region is used, leading to instability in local estimates, especially for small regions. We show that borrowing information across regions using a Bayesian hierarchical model stabilizes estimates, leading to lower variability and improved predictive utility. Specifically, our approach more effectively identifies regions enriched for ClinVar pathogenic variants. We also identify significant correlations between sub-region intolerance and the distribution of pathogenic variation in disease-associated genes, with AUCs for classifying de novo missense variants in Online Mendelian Inheritance in Man (OMIM) genes of up to 0.86 using exonic sub-regions and 0.91 using sub-regions defined by protein domains. This result immediately suggests that considering the intolerance of regions in which variants are found may improve diagnostic interpretation. We also illustrate the utility of integrating regional intolerance into gene-level disease association tests with a study of known disease-associated genes for epileptic encephalopathy.

Integrating predicted transcriptome from multiple tissues improves association detection.

Integration of genome-wide association studies (GWAS) and expression quantitative trait loci (eQTL) studies is needed to improve our understanding of the biological mechanisms underlying GWAS hits, and our ability to identify therapeutic targets. Gene-level association methods such as PrediXcan can prioritize candidate targets. However, limited eQTL sample sizes and absence of relevant developmental and disease context restrict our ability to detect associations. Here we propose an efficient statistical method (MultiXcan) that leverages the substantial sharing of eQTLs across tissues and contexts to improve our ability to identify potential target genes. MultiXcan integrates evidence across multiple panels using multivariate regression, which naturally takes into account the correlation structure. We apply our method to simulated and real traits from the UK Biobank and show that, in realistic settings, we can detect a larger set of significantly associated genes than using each panel separately. To improve applicability, we developed a summary result-based extension called S-MultiXcan, which we show yields highly concordant results with the individual level version when LD is well matched. Our multivariate model-based approach allowed us to use the individual level results as a gold standard to calibrate and develop a robust implementation of the summary-based extension. Results from our analysis as well as software and necessary resources to apply our method are publicly available.