Robust and annotation-free analysis of alternative splicing using short-read scRNA-seq data

Although alternative splicing is a fundamental and pervasive aspect of gene expression in higher eukaryotes, it is often omitted from single-cell studies due to quantification challenges inherent to commonly used short-read sequencing technologies. Here, we propose a new computational method, scQuint, that performs accurate quantification, dimensionality reduction, and differential splicing analysis using full-length, short-read, single-cell RNA-seq data. scQuint does not require transcriptome annotations and is robust to technical artifacts. In applications across diverse mouse tissues from Tabula Muris and the primary motor cortex from the BRAIN Initiative Cell Census Network, we find evidence of strong cell-type-specific alternative splicing, complementary to total gene expression, and identify a large volume of previously unannotated splice junctions. To further elucidate the regulation of alternative splicing, we build a predictive model based on splicing factor activity, which recovers several known interactions and generates new hypotheses, including potential regulatory roles of novel alternative splicing events in Khdrbs3 and Rbfox1.