Supervised Classification Approach for Precise Cell Type Identification Improves Single Cell Data Analysis
摘要
Single Cell RNA sequencing technology aims towards addressing the problem of cellular heterogeneity in biological systems. It unveils the transcriptomic differences at a single cell level within a cell population. Precise cell type prediction within apparently similar cells with subtle differences among them greatly influences single cell RNA sequence data analysis. The proposed machine learning based prediction model attempts to classify different cellular subtypes within apparently similar cell population. The prediction model has been trained with 68k Peripheral Blood Mononuclear Cell (PBMC) dataset. It predicted cellular subtypes on an unknown Peripheral Blood Mononuclear Cell 4k validation dataset with greater accuracy.