Computational identification of splicing phenotypes from single cell transcriptomic experiments

Abstract

RNA splicing is an important driver of heterogeneity in single cells, both through the expression of alternative transcripts and as a major determinant of transcriptional kinetics. However, the intrinsic coverage limitations of scRNA-seq technologies make it challenging to associate specific splicing events to cell-level phenotypes. Here, we present BRIE2, a scalable computational method that resolves these issues by regressing single-cell transcriptomic data against cell-level features. We show that BRIE2 effectively identifies differential alternative splicing events that are associated with a disease. Additionally, BRIE2 allows a principled selection of genes (differential momentum genes) that capture heterogeneity in transcriptional kinetics and improve quantitatively RNA velocity analyses. BRIE2, therefore, extends the scope of single-cell transcriptomic experiments towards the identification of splicing phenotypes associated with biological changes at the single-cell level.