A contamination focused approach for optimizing thesingle-cell RNA-seq experiment

Abstract

Achieving high data quality in single-cell RNA-seq (scRNA-seq) experiments has always been a significant challenge stemming from minute signal that can be detected in individual cells. Droplet-based scRNA-seq additionally suffers from ambient contamination, comprising nucleic acid materials released by dead cells into the loading buffer and co-encapsulated with real cells, which further washes out real biological signals. Here, we developed quantitative, ambient contamination-based metrics and an associated software package that can both evaluate current datasets and guide new experimental optimizations. We performed a series of experimental optimizations using the inDrops platform to address the mechanical and microfluidic cell encapsulation aspect of an scRNA-seq experiment, with a focus on minimizing ambient contamination. We report improvements that can be achieved via cell fixation, microfluidic loading, microfluidic dilution, and nuclei versus cell preparation; many of these parameters are inaccessible on commercial platforms. We provide insights into previously obscured factors that can affect scRNA-seq data quality and suggest mitigation strategies that can guide future experiments.