Conbase: a software for discovery of clonal somatic mutations in single cells through read phasing

Somatic variant analysis in single cells is challenging due to the large fraction of false positive and false negative variant calls. To overcome these obstacles we developed Conbase, a software for calling clonal somatic mutations in genome sequencing data from single cells. Conbase takes advantage of read phasing and observations across the single cell dataset and an unamplified bulk sample, to mitigate effects of amplification errors and other aspects of bioinformatic analysis, including alignment artefacts and an incomplete reference genome with respect to the genome of the donor. This strategy enables determination of presence or absence of mutations despite low read depth and high rates of allelic dropout (ADO) (>90%). We demonstrate the potential of Conbase by unambiguously defining the clonal relationships of single cells from two in vivo expanded T cell clones isolated directly from the peripheral blood of a healthy human donor.