STANDARDIZE QC PROCEDURE FOR SCRNA-SEQ

Abstract

Single-cell RNA sequencing (scRNA-seq) has emerged as a pivotal technology for dissecting cellular heterogeneity and function. In an effort to assess the consistency and rigor of quality control (QC) measures across scRNA-seq studies, we systematically reviewed publications from high-impact journals, including Cell, Nature, Science, and their major sister journals. Our analysis revealed a lack of standardization in QC procedures, with significant variability in the parameters employed. Despite general agreement on the necessity of certain QC steps, such as the removal of low-quality cells and the detection of doublets, the specific criteria for these steps were often arbitrarily defined and not universally applied. Notably, the assessment of ambient RNA contamination and the precision of gene expression measurements were frequently overlooked, potentially leading to the inclusion of spurious data in downstream analyses. To address these inconsistencies, we propose a revised set of QC procedures and parameters, which yielded distinct results compared to the original publications when applied to existing datasets. Moreover, we also assessed the performance of the existing data-driven QC tools in distinguishing the low-quality cells from the high-quality cells. Our findings underscore the urgent need for a standardized approach to QC in scRNA-seq to ensure the reliability and reproducibility of biological insights derived from this powerful technology.