IRescue: uncertainty-aware quantification of transposable elements expression at single cell level

Transposable elements (TEs) are mobile DNA repeats that contribute to the evolution of eukaryotic genomes. In complex organisms, TE expression is tissue specific. However, their contribution to cellular heterogeneity is still unknown and challenging to investigate in single-cell RNA sequencing (scRNA-seq), due to the ubiquity and homology of TEs in the genome. We introduce IRescue (Interspersed Repeats single-cell quantifier), the first software that accurately estimates the expression of TE subfamilies at single-cell level, implementing a UMI deduplication algorithm to allocate reads ambiguously mapped on TEs, while correcting for UMI sequencing errors. Applying IRescue on simulated datasets and real scRNA-seq of colorectal cancers, we could precisely estimate TE subfamilies expression. We show that IRescue improves the definition of cellular heterogeneity, detecting TE expression signatures and specific TE-containing splicing isoforms.