Identification and analysis of splicing quantitative trait loci across multiple tissues in the human genome

Abstract

We have developed an efficient and reproducible pipeline for the discovery of genetic variants affecting splicing (sQTLs), based on an approach that captures the intrinsically multivariate nature of this phenomenon. We employed it to analyze the multi-tissue transcriptome GTEx dataset, generating a comprehensive catalogue of sQTLs in the human genome. A core set of these sQTLs is shared across multiple tissues. Downstream analyses of this catalogue contribute to the understanding of the mechanisms underlying splicing regulation. We found that sQTLs often target the global splicing pattern of genes, rather than individual splicing events. Many of them also affect gene expression, but not always of the same gene, potentially uncovering regulatory loci that act on different genes through different mechanisms. sQTLs tend to be preferentially located in introns that are post-transcriptionally spliced, which would act as hotspots for splicing regulation. While many variants affect splicing patterns by directly altering the sequence of splice sites, many more modify the binding of RNA-binding proteins (RBPs) to target sequences within the transcripts. Genetic variants affecting splicing can have a phenotypic impact comparable or even stronger than variants affecting expression, with those that alter RBP binding playing a prominent role in disease. 33 splice sites, many more modify the binding of RNA-binding proteins (RBPs) to target sequences within 34 the transcripts. We have observed that sQTLs often impact GWAS traits and diseases more than variants 35 affecting only gene expression, confirming earlier reports which suggest that splicing mutations underlie 36 many hereditary diseases 15,19 . For many conditions, GWAS associations are particularly strong for sQTLs 37 3 altering RBP binding sites. 40 For sQTL mapping, we developed sQTLseekeR2, a software based on sQTLseekeR 20 , which identi-41 fies genetic variants associated with changes in the relative abundances of the transcript isoforms of a 42 given gene. sQTLseekeR uses the Hellinger distance to estimate the variability of isoform abundances 43 across observations, and Anderson's method 21,22 , a non-parametric analogue to multivariate analysis 44 of variance, to assess the significance of the associations (see Methods and Supplementary Note 1). 45 Among other enhancements, sQTLseekeR2 improves the accuracy and speed of the p value calcula-46 tion, and allows to account for additional covariates before testing for association with the genotype, 47 while maintaining the multivariate statistical test in sQTLseekeR. It also implements a multiple testing 48 correction scheme that empirically characterizes, for each gene, the distribution of p values expected un-49 der the null hypothesis of no association (see Methods and Supplementary Note 1). To ensure highly 50 parallel, portable and reproducible sQTL mapping, we embedded sQTLseekeR2 in a Nextflow 23 (plus 51 Docker, https://www.docker.com/) computational workflow named sqtlseeker2-nf, available at 52 https://gi...

Alternative splicing: An overlooked mechanism contributing to local adaptation?

“…DS, but not DE, genes were also associated with more GO terms than nondifferentiated genes, suggesting that DS genes may have greater pleiotropy than DE genes. These data match findings from tissue‐specific splicing QTLs (sQTLs) in humans, which were found to be more pleiotropic and have a stronger effect on phenotype than expression QTLs (eQTLs) (Garrido‐Martín et al, 2021). Thus, identifying sQTL may be necessary to fully understand the molecular bases of adaptation.…”

Clinical implementation of RNA sequencing for Mendelian disease diagnostics

“…This is the case for NMD (90). Also, a minor proportion of variants found in GTEx associated with splicing show tissue-specific effects (96).…”

The EN-TEx resource of multi-tissue personal epigenomes & variant-impact models

“…S6.1e shows more detail on the splicing isoforms. The differences in splicing between individuals 2 and 3 could reflect the SV disrupting regions important to splicing, as suggested by the known GTEx sQTL sites nearby (72). Alternatively, the differences in splicing between the two samples may be caused by tissue specificity: the individual 2 sample is from the adrenal gland, and the individual 3 sample is from the heart left ventricle.…”