Alternative polyadenylation (APA) is a ubiquitous post-transcriptional regulatory mechanism. This process has been studied using bulk-RNA sequencing, and has been demonstrated to be cell-type specific and associated with cancer, development and differentiation.

With the advent of single-cell RNA sequencing, we can probe into the cellular transcription at greater resolution. The oligo-dT capturing protocol naturally lends itself to studying APA events, while there is also strong interest to investigate APA on a single-cell level, bypassing tissue heterogeneity and allows us to interrogate subsets of cells in different states.

We present a novel method to analyse APA on a single-cell level, with a pipeline to extract supporting reads, to determine and rank transcript end sites, and to a...[ Read more ]

Alternative polyadenylation (APA) is a ubiquitous post-transcriptional regulatory mechanism. This process has been studied using bulk-RNA sequencing, and has been demonstrated to be cell-type specific and associated with cancer, development and differentiation.

With the advent of single-cell RNA sequencing, we can probe into the cellular transcription at greater resolution. The oligo-dT capturing protocol naturally lends itself to studying APA events, while there is also strong interest to investigate APA on a single-cell level, bypassing tissue heterogeneity and allows us to interrogate subsets of cells in different states.

We present a novel method to analyse APA on a single-cell level, with a pipeline to extract supporting reads, to determine and rank transcript end sites, and to annotate APA events to individual cells based on the output.

The method is then applied to the Chinese Glioma Genome Atlas dataset, performing subsequent analysis to integrate the APA data with existing single-cell annotations. We identified a global lengthening trend in APA isoforms among cancer cells together with APA candidates associated with cancer process. We also characterized the single-cell APA stochasticity, and outlined a network of APA event-regulator interactions.

This method performs adequately in 10X sequencing protocols, and could provide complementary information in future single-cell research.