Quantification of transcript isoforms at the single-cell level using SCALPEL

<p>Single-cell RNA sequencing (scRNA-seq) has facilitated the study of gene  expression and the development of new tools to quantify transcript in  individual cells. Yet, most of these methods have low sensitivity and  accuracy. Here we present SCALPEL, a Nextflow-based tool to quantify and  characterize transcript isoforms at the single-cell level using  standard 3’ based scRNA-seq data. Using synthetic data, we show that  SCALPEL predictions have higher sensitivity and specificity than other  tools and can be validated experimentally. Using real datasets, SCALPEL  predictions have a high agreement with that of other tools in both 10x  and Drop-seq datasets. We have used SCALPEL to study the changes in  isoform usage during mouse spermatogenesis and in the differentiation of  induced pluripotent stem cells (iPSCs) to neural progenitors. These  analyses allow the identification of novel cell populations that cannot  be defined using conventional gene expression profiles, confirm known  changes in 3’ UTR length during cell differentiation, and identify  cell-type specific miRNA signatures controlling isoform expression in  individual cells. Additionally, we showcase the value of SCALPEL to  improve isoform quantification using paired long and short scRNA-seq  data. Together, our work highlights how SCALPEL expands the current  scRNA-seq toolset to explore post-transcriptional gene regulation in  individual cells from different species, tissues, and technologies to  investigate the variability and the specificity of gene regulatory  mechanisms at the single-cell level.</p>