We conducted a comprehensive study of copy number variants [CNVs] well-tagged by SNPs [r2≥0.8] by analyzing their effect on gene expression and their association with disease susceptibility and other complex human traits. We tested whether these CNVs were more likely to be functional than frequency-matched SNPs as trait-associated loci or as expression quantitative trait loci [eQTLs] influencing phenotype by altering gene regulation. Our study found that CNV–tagging SNPs are significantly enriched for cis eQTLs; furthermore, we observed that trait associations from the NHGRI catalog show an overrepresentation of SNPs tagging CNVs relative to frequency-matched SNPs. We found that these SNPs tagging CNVs are more likely to affect multiple expression traits than frequency-matched variants. Given these findings on the functional relevance of CNVs, we created an online resource of expression-associated CNVs [eCNVs] using the most comprehensive population-based map of CNVs to inform future studies of complex traits. Although previous studies of common CNVs that can be typed on existing platforms and/or interrogated by SNPs in genome-wide association studies concluded that such CNVs appear unlikely to have a major role in the genetic basis of several complex diseases examined, our findings indicate that it would be premature to dismiss the possibility that even common CNVs may contribute to complex phenotypes and at least some common diseases.
Author Summary
Despite the large number of SNPs found to be reproducibly associated with complex diseases, they collectively account for only a small proportion of the overall heritability to such traits. CNVs have thus been proposed to explain some of the missing heritability and to alter disease susceptibility. However, a recent study of the genetics of 8 common diseases involving 16,000 cases and 3,000 controls failed to identify any novel CNVs associated with disease and concluded that CNVs are unlikely to play a major role in their etiology. Studies we report here show that we must be careful not to dismiss the possibility that CNVs may indeed underlie some of the observed associations with complex disease. Our findings show that well-tagged CNVs are disproportionately more likely to be eQTLs, as well as cis-eQTLs, than frequency-matched SNPs; furthermore, reproducible trait associations, as represented in the NHGRI catalog, are enriched for well-tagged CNVs than frequency-matched SNPs. Because of these findings on the strong functional relevance of these CNVs, we created a database [available at //www.scandb.org/] of expression associated CNVs to supplement our earlier studies of SNP eQTLs and to contribute to future studies of the genetics of complex traits.
Citation: Gamazon ER, Nicolae DL, Cox NJ [2011] A Study of CNVs As Trait-Associated Polymorphisms and As Expression Quantitative Trait Loci. PLoS Genet 7[2]: e1001292. //doi.org/10.1371/journal.pgen.1001292
Editor: Vivian G. Cheung, University of Pennsylvania, United States of America
Received: August 15, 2010; Accepted: January 5, 2011; Published: February 3, 2011
Copyright: © 2011 Gamazon et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This work was funded through PAAR [Pharmacogenetics of Anti-Cancer Agents Research; U01 GM61393], ENDGAMe [ENhancing Development of Genome-Wide Association Methods] initiative [U01 HL084715], the Genotype-Tissue Expression project [GTeX] [R01 MH090937], The University of Chicago Breast Cancer SPORE [P50 CA125183], and The University Of Chicago DRTC [Diabetes Research and Training Center; P60 DK20595]. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Although genome wide association studies [GWAS] have considerably advanced our knowledge of the genetics of complex traits, they fail to account for the bulk of the overall heritability . Structural variation, including copy number variants [CNVs], may account for some of the missing heritability, but a comprehensive study of the contribution of these variants to complex traits genetics is generally lacking. One of the main challenges of integrating CNVs into GWAS is the reliability of their assay, but improvements in genotyping platforms and the development of algorithms for the analysis of CNVs increasingly facilitate the investigation of the role of CNVs in disease susceptibility . Given the effect of adjacent CNVs on SNP intensity data, early SNP arrays were designed with the consequence that assays from many CNV loci were excluded, but a newer generation of arrays have been implemented that combine SNP assays and copy number probes optimized for copy-number measurement for target regions of known CNVs . New algorithms are also facilitating higher resolution maps of common CNVs [those that segregate at an allele frequency >5%].
Partly due to these technological and analytic developments, the search for links between CNVs and disease susceptibility has been actively pursued in recent years. Structural variation has of course long been known to influence such Mendelian disorders as Williams-Beuren syndrome or Charcot-Marie Tooth neuropathy Type 1A, but more recently there has been widespread interest in the study of CNVs as mediators of more common complex diseases. The contribution of structural variation to certain neurodevelopmental disorders such as schizophrenia and autism spectrum disorder has been investigated by recent studies of de novo copy number variation . A common 20 kb deletion upstream of IRGM in perfect linkage disequilibrium [r2 = 1.0] with the SNP most strongly associated with Crohn's disease in the region has recently been shown to alter IRGM regulation, which in turn is known to affect autophagy, consistent with the deletion polymorphism being the causal allele . Similarly, a 45 kb deletion polymorphism in NEGR1 has been shown to be the likely causal variant for the reproducible association with body mass index in humans, demonstrating a neuronal influence on body weight regulation . These discoveries and similar developments suggest that the interpretation of known disease associations continues to be fraught with challenges, as CNVs may be in linkage disequilibrium [LD] with associated variants, and that the identification of causal variants requires an approach that considers both SNPs and CNVs at associated loci.
The recent release of a reference map of human CNVs in the HapMap populations promises to shed light on the role of these variants in disease pathogenesis. On the other hand, a recent comprehensive study involving 16,000 cases of eight common diseases and 3000 shared controls strongly suggests that CNVs are likely to make a relatively minor contribution to disease susceptibility. In the present study, we approached the investigation of this important topic by exploring evidence of enrichment of CNVs among trait-associated loci in disease susceptibility and in disease classes, as represented by the NHGRI catalog of reproducible associations from genome-wide association scans , for certain classes of CNVs, and by conducting a comprehensive functional characterization of CNVs as expression quantitative trait loci [eQTLs]. The former seeks to evaluate the genetic information contained in CNVs in regard to disease risk and other complex traits; the latter assesses the functional impact of CNVs on the transcriptome, through which they may contribute to or cause disease phenotypes and other complex traits .
Results
Tagged CNVs [tCNVs] are enriched for eQTLs and cis-acting eQTLs
summarizes the CNVs included in our study. Because of the ease with which SNPs can be utilized in enrichment studies appropriately conditioned on minor allele frequency, we utilized tCNVs [CNVs tagged by SNPs] for many of these studies, but did not otherwise consider them to be a special class of CNVs.
Download:
PPT
PowerPoint slide
PNG
larger image
TIFF
original image
Table 1. CNV categories that were investigated for role in transcriptional expression.
//doi.org/10.1371/journal.pgen.1001292.t001
At an expression p value threshold of 10−4, tCNVs from the recently released study [N = 3,432] of CNVs in 16,000 cases and 3,000 controls were evaluated for their effect on gene expression. We first considered tCNVs that are tagged at r2≥0.80. The tag SNPs for these CNVs enable SNP-based analyses and simulation studies. We identified 1,714 such SNPs tagging CNVs [corresponding to 1,761 tCNVs], of which 532 are eQTLs at this threshold; collectively, these SNPs tagging CNVs were found to target 2,215 distinct transcripts. Restricting our focus to eQTLs that regulate transcript levels in cis [defined as within 4 MB of target transcript], we observed a highly significant cis eQTL enrichment [p