MitoSort: robust demultiplexing of pooled single-cell genomics data using endogenous germline mitochondrial variants

Abstract

Multiplexing across donors has emerged as a popular strategy to increase throughput, reduce costs, overcome technical batch effects, and improve doublet detection in single-cell genomic studies. Using endogenous genetic barcodes eliminates the need for additional experimental processing steps. Among the available choices for endogenous barcodes, the unique features of mtDNA variants render them a more computationally efficient and robust option compared to genome variants. Here we present MitoSort, a method that uses mtDNA germline variants to assign cells to their donor of origin and identify cross-genotype doublets. We evaluated the performance of MitoSort by in silico pooled mtscATAC-seq libraries and experimentally multiplexed data using cell hashing method. MitoSort achieve both high accuracy and efficiency on genotype clustering and doublet detection for mtscATAC-seq data, which fills a void left by the inadequacies of current computational techniques tailored for scRNA-seq data. Moreover, MitoSort exhibits versatility and can be applied to various single-cell sequencing approaches beyond mtscATAC-seq, as long as the mtDNA variants can be reliably detected. Furthermore, through a case study, we demonstrated that demultiplexing 8 individuals assayed at the same time with MitoSort, enables the comparison of cell composition without batch effects.