Density peaks based clustering for single-cell interpretation via multikernel learning

The development in single-cell technology has enabled to quantify the high throughput gene expressions of individual cell, and it became possible to discover heterogeneity at cell level. To detect heterogeneity within cell population remains challenging in presence of outliers, biological noise, and dropouts. SIMLR (single-cell interpretation via multikernel learning) has been proposed to measure cell to cell similarity, dimensional reduction, clustering, and visualization of scRNA-seq data. SIMLR uses K-means to organize the cells into the predefined number of types, which is a significant drawback of SIMLR toward adaptive analysis of scRNA-seq data. In this paper, we introduced density peaks based clustering for single-cell interpretation via multikernel learning (DP-SIMLR), an adaptive approach to discover biological meaningful heterogeneity within the individual cell population. The DP-SIMLR is an extension of SIMLR, where the concept of density peaks is employed to discover heterogeneity within the cell population, adaptively. We have evaluated the DP-SIMLR on four scRNA-seq datasets and the results are compared with SIMLR.