Systematic comparative analysis of single cell RNA-sequencing methods

Abstract

A multitude of single-cell RNA sequencing methods have been developed in recent years, with dramatic advances in scale and power, and enabling major discoveries and large scale cell mapping efforts. However, these methods have not been systematically and comprehensively benchmarked. Here, we directly compare seven methods for single cell and/or single nucleus profiling from three types of samples -cell lines, peripheral blood mononuclear cells and brain tissue -generating 36 libraries in six separate experiments in a single center. To analyze these datasets, we developed and applied scumi, a flexible computational pipeline that can be used for any scRNA-seq method. We evaluated the methods for both basic performance and for their ability to recover known biological information in the samples. Our study will help guide experiments with the methods in this study as well as serve as a benchmark for future studies and for computational algorithm development.

scClassify: hierarchical classification of cells

“…2c, Supplementary Fig 5) as query, achieving a high 86 concordance (~85% accuracy). Similar results were found ( Supplementary Fig 6, 7) for three mouse 87 neuronal datasets 9,10 , each with about 20 cell types. 88 89 scClassify labels cells from a query dataset "unassigned" when their type is not in the reference 90 dataset.…”

“…The PBMC data collection [9] was downloaded from https:// portals.broadinstitute.org/single_cell/study/SCP424/ single-cell-comparison-pbmc-data. It contains a collection of seven datasets that were sequenced using different platforms (Smart-seq, Cel-seq, inDrops, dropSeqs, se-qWells, 10X Genomics (V3), 10X Genomics (V2))).…”

“…We used scMerge [21] to remove the differences between two batches in the PBMC data collection [9]. PBMC dataset generated from two samples using five different protocols and then construct the learning curve.…”

“…We used scMerge [21] to remove batch effects between two samples in the PBMC data collection generated from five different protocols [9] and next constructed the learning curves. We found that in most of the cases (SMART-seq, CEL-seq, and inDrops), the learning curve based on batch-corrected data achieved a higher accuracy rate that the uncorrected data (Supplementary Fig.…”

“…To test scClassify with a 85 large dataset having a complex cell type hierarchy, we used the Tabula Muris FACS dataset 8 as 86 reference and the Microfluidic dataset ( Fig. 2c 9,10 , each with about 20 cell types. 89 90 scClassify labels cells from a query dataset "unassigned" when their type is not in the reference 91 dataset.…”