Linnorm: Improved statistical analysis for single cell RNA-seq expression data

Abstract

Linnorm is an R package for the analysis of single cell RNA sequencing (scRNA-seq) data based on a novel normalizing transformation method. Previous works on scRNA-seq analysis methods have focused on downstream analysis methods by utilizing the conventional relative expression units and/or the log plus one transformation. Moreover, RNA-seq normalization methods are still commonly used for scRNA-seq. We present a scRNA-seq data oriented normalization and transformation method. It allows precise normalization and transformation by filtering of the dataset with or without spike-ins. Our assessments showed that Linnorm performs better than existing methods (edgeR, DESeq2, voom, Seurat etc) in terms of false positive rate control, differential gene expression analysis, clustering analysis and speed. Moreover, we show that existing methods can benefit from Linnorm’s normalization and transformation methods