Double-jeopardy: scRNA-seq doublet/multiplet detection using multi-omic profiling

Abstract

Summary: The computational detection and exclusion of cellular doublets and/or multiplets is a cornerstone for the identification the true biological signals from single-cell RNA sequencing (scRNA-seq) data. Current methods do not sensitively identify both heterotypic and homotypic doublets and/or multiplets. Here, we describe a machine learning approach for doublet/multiplet detection utilizing VDJ-seq and/or CITE-seq data to predict their presence based on transcriptional features associated with identified hybrid droplets. This approach highlights the utility of leveraging multi-omic single-cell information for the generation of high-quality datasets. Our method has high sensitivity and specificity in inflammatory-cell-dominant scRNA-seq samples, thus presenting a powerful approach to ensuring high-quality scRNA-seq data. Motivation: Single-cell RNA sequencing (scRNA-seq) techniques are transforming our understanding of multicellular organisms, disease states, and cellular heterogeneity. The aggregation of two or more cells into single droplets (doublets/multiplets) during the cell capture step of scRNA-seq resulting in hybrid transcriptomes can lead to false discoveries of rare cell types, intermediate cell states, and disease-associated transcriptomic signatures. Current methods do not sensitively identify many doublets/multiplets. Here, we address this doublet/multiplet detection issue utilizing VDJ-seq and/or CITE-seq, and apply machine learning to predict their presence based on transcriptional features associated with identified hybrid droplets, and ultimately scRNA-seq data quality.

Keywords

• single-cell transcriptomics
• B cell receptor
• T cell receptor
• doublets
• multi-omics profiling
• ADT