Searching match for single-cell open-chromatin profiles in large pools of single-cell transcriptomes and epigenomes for reference supported analysis

Abstract

The true benefits of large datasets of the single-cell transcriptome and epigenome profiles can be availed only with their inclusion and search for annotating individual cells. Matching a single cell epigenome profile to a large pool of reference cells remains a major challenge. We developed a method (scEpiSearch) to resolve the challenges of searching and comparing single-cell open-chromatin profiles against large pools of single-cell expression and open chromatin datasets. scEpiSearch is more accurate than other methods when comparing single cell open-chromatin profiles to single-cell transcriptomes and epigenomes. scEpiSearch also provides a robust method for reference-supported co-embedding of single-cell open chromatin profiles. In performance benchmarks, scEpiSearch outperformed multiple methods for the low dimensional co-embedding of single-cell open-chromatin profiles irrespective of platforms and species. scEpiSearch works with both reference single-cell expression and epigenome profiles, enabling classification of single-cell open-chromatin profiles. Here we demonstrate the unconventional utilities of scEpiSearch by applying it on single-cell epigenome profiles of K562 cells and samples from patients with acute leukaemia to reveal different aspects of their heterogeneity, multipotent behaviour and de-differentiated states. Applying scEpiSearch on our single-cell open-chromatin profiles from embryonic stem cells(ESCs), we identified ESC subpopulations with more activity and poising for endoplasmic reticulum stress and unfolded protein response. Thus, scEpiSearch solves the non-trivial problem of amalgamating information from a large pool of single-cells to identify and study the regulatory states of cells using their single-cell epigenomes.The true benefits of large datasets of the single-cell transcriptome and epigenome profiles can be availed only with their inclusion and search for annotating individual cells.