A powerful and flexible statistical framework for testing hypotheses of allele-specific gene expression from RNA-seq data

Abstract

Variation in gene expression is thought to make a significant contribution to phenotypic diversity among individuals within populations. Although high-throughput cDNA sequencing offers a unique opportunity to delineate the genomewide architecture of regulatory variation, new statistical methods need to be developed to capitalize on the wealth of information contained in RNA-seq data sets. To this end, we developed a powerful and flexible hierarchical Bayesian model that combines information across loci to allow both global and locus-specific inferences about allele-specific expression (ASE). We applied our methodology to a large RNA-seq data set obtained in a diploid hybrid of two diverse Saccharomyces cerevisiae strains, as well as to RNA-seq data from an individual human genome. Our statistical framework accurately quantifies levels of ASE with specified false-discovery rates, achieving high reproducibility between independent sequencing platforms. We pinpoint loci that show unusual and biologically interesting patterns of ASE, including allele-specific alternative splicing and transcription termination sites. Our methodology provides a rigorous, quantitative, and high-resolution tool for profiling ASE across whole genomes. [Supplemental material is available for this article.]Gene expression is the fundamental initial step in the process by which static genomic information gives rise to dynamic organismal phenotypes. Variation in gene expression has the potential to contribute significantly to phenotypic diversity within species and divergence between species (Britten and Davidson 1971;King and Wilson 1975). There is a diverse array of well-characterized examples of phenotypes influenced by regulatory polymorphisms, ranging from pelvic morphology in sticklebacks (Chan et al. 2010) to malaria susceptibility in humans (Tournamille et al. 1995). Although heritable variation in gene expression levels appears to be ubiquitous among individuals within species, an understanding of the distribution of regulatory variation and the mechanisms by which regulatory polymorphisms act remains limited (Rockman and Kruglyak 2006;Skelly et al. 2009).Heritable differences in gene expression between individuals are ultimately caused by polymorphisms that affect the expression level of either one or both alleles (cis-acting or trans-acting polymorphisms, respectively) in a diploid. A powerful approach for identifying cis-acting regulatory variation is measuring allelespecific expression (ASE). An observation of differential allelic expression in a heterozygote indicates the presence of one or more variants that act in cis to affect the expression level of the gene. (Knight et al. 2003). A unifying theme of these studies has been that ASE is widespread both within and between a wide variety of species.More recently, high-throughput sequencing has been used to assess ASE (Degner et al. 2009;Main et al. 2009;Zhang et al. 2009;Emerson et al. 2010;Heap et al. 2010;McManus et al. 2010;Montgomery et al. 2010;Pickrell et al. 2010), whic...

SNP-guided identification of monoallelic DNA-methylation events from enrichment-based sequencing data

“…After pre-processing the data for each variant site, binomial tests were performed. Using an FDR of 1%, and requiring the presence of at least two significant variant sites per gene, in total, 19 840 genes showed ASE, ∼1190 genes per tissue (Supplementary Table S5), in line with the amount of loci identified in previous studies (1,(41)(42)(43). In a next step, it was examined if some genes from Table 2--the genes with one (or more) monoallelic methylated SNP(s) in their genic regions--were also characterized by ASE.…”

Differential principal component analysis of ChIP-seq

“…ChIP-seq provides new opportunities to study allele-specific binding (ASB) and HM (22)(23)(24). ASB detection often suffers from low statistical power because only reads mapped to heterozygote SNPs contain allelic information.…”

Allelic mapping bias in RNA-sequencing is not a major confounder in eQTL studies

“…Similar problems may arise in alignment of ChIP-Seq reads and chromatin state QTL mapping. In RNA-seq analysis, accounting for genotypedependent mapping bias has been important for obtaining more accurate and reliable results from analysis of allelic specific expression (ASE) [8,[11][12][13][14], allelic specific binding (ASB) [12], and DNaseI sensitivity QTLs [15]; nevertheless, similar analyses have not been done for eQTLs.…”