Vladimir Brusic
University of Nottingham Ningbo China
Abstract
Single cell transcriptomics is a rapidly growing area with an urgent need for new analytical tools to complement and supersede unsupervised clustering. We defined a new method for deriving gene expression profiles from single-cell gene expression matrices. We named these profiles the “single-cell-derived-class” (SCDC) profiles. We developed SCDC profiles for multiple cell types and subtypes of peripheral blood mononuclear cells (PBMC) using the results of single cell transcriptomics (SCT) experiments. SCDC profiles represent characteristic patterns of gene expressions of the types and subtypes of healthy human PBMC. We studied the reproducibility of SCDC profiles, their robustness, and their applications in classifying healthy human PBMC types and subtypes. SCDC profiles are efficient and convenient tools for the analysis of SCT data derived from PBMC samples. These profiles are highly reproducible, even when derived from unrelated studies, provided that the sample processing steps are comparable and the same SCT technology is used. The classification accuracy of SCDC profiles is high. SCDC profiles can be used for supervised classification and the discovery of new subtypes of PBMC.
