Multivariate analysis, Data visualization, Alternative splicing, Quantitative trait loci, Genome-wide association studies, Análisis multivariante, Visualización de datos, Splicing alternativo, Loci de caracteres cuantitativos, and Estudios de asociación del genoma completo
We have developed an efficient and reproducible pipeline for the identification of genetic variants affecting splicing (splicing quantitative trait loci or sQTLs), based on an approach that captures the intrinsically multivariate nature of this phenomenon. We employed it to study the multi-tissue transcriptome GTEx dataset, generating a comprehensive catalogue of sQTLs in the human genome. Downstream analyses of this catalogue provide novel insights into the mechanisms underlying alternative splicing regulation and its contribution to human complex traits and diseases. To facilitate the visualization of splicing events in GTEx and other large-scale RNA-seq studies, we developed a software to generate sashimi plots, which supports the aggregated representation of hundreds of samples. Given the growing interest in efficient methods to identify genetic effects on multiple traits, we extended the statistical framework employed for sQTL mapping (Anderson test) to accommodate any quantitative multivariate phenotype and experimental design. We derived the limiting distribution of the test statistic, allowing to compute asymptotic p values. We further demonstrated the advantages and applicability of our approach to GWAS and QTL mapping analyses using simulated and real datasets.