Abstract:
|
In recent years, single cell RNA-Seq technology has gained in popularity due to its ability to study cell to cell heterogeneity and the detection of novel cell types. Many methods have been developed for differentially expression of genes (DEG) analysis in single cells including traditional bulk RNA-seq methods (edgeR, limma-voom, and DESeq2) and single cell specific DEG methods (Seurat). In this work, we extend previous work on the comparison of DEG analysis methods of single cell data to the 10x genomics platform. More specifically, we compared 7 DEG analysis methods on 10x genomics data from iPSC motor neurons, dopaminergic neurons, and synthetic datasets for the metrics of type-1, false discovery rate, power, ROC-auc, and computation time. Additionally, we explored multiple filtering scenarios and their impact on DEG performance.
|