RT Journal
A1 Schlattl, Andreas
A1 Anders, Simon
A1 Waszak, Sebastian M.
A1 Huber, Wolfgang
A1 Korbel, Jan O.
T1 Relating CNVs to transcriptome data at fine resolution: Assessment of the effect of variant size, type, and overlap with functional regions
JF Genome Research 
JO Genome Research 
YR 2011 
FD December 01 
VO 21 
IS 12 
SP 2004 
OP 2013 
DO 10.1101/gr.122614.111 
UL http://genome.cshlp.org/content/21/12/2004.abstract 
AB Copy-number variants (CNVs) form an abundant class of genetic variation with a presumed widespread impact on individual traits. While recent advances, such as the population-scale sequencing of human genomes, facilitated the fine-scale mapping of CNVs, the phenotypic impact of most of these CNVs remains unclear. By relating copy-number genotypes to transcriptome sequencing data, we have evaluated the impact of CNVs, mapped at fine scale, on gene expression. Based on data from 129 individuals with ancestry from two populations, we identified CNVs associated with the expression of 110 genes, with 13% of the associations involving complex, multiallelic CNVs. Categorization of CNVs according to variant type, size, and gene overlap enabled us to examine the impact of different CNV classes on expression variation. While many small (<4 kb) CNVs were associated with expression variation, overall we observed an enrichment of large duplications and deletions, including large intergenic CNVs, relative to the entire set of expression-associated CNVs. Furthermore, the copy number of genes intersecting with CNVs typically correlated positively with the genes' expression, and also was more strongly correlated with expression than nearby single nucleotide polymorphisms, suggesting a frequent causal role of CNVs in expression quantitative trait loci (eQTLs). We also elucidated unexpected cases of negative correlations between copy number and expression by assessing the CNVs' effects on the structure and regulation of genes. Finally, we examined dosage compensation of transcript levels. Our results suggest that association studies can gain in resolution and power by including fine-scale CNV information, such as those obtained from population-scale sequencing.