Multiplexed droplet single-cell RNA-sequencing using natural genetic variation

Area of Science:

• Genomics
• Computational Biology
• Molecular Biology

Background:

• Droplet single-cell RNA sequencing (dscRNA-seq) offers high-throughput transcriptome profiling.
• Challenges in dscRNA-seq include inefficient sample processing and technical batch effects, hindering multi-individual analyses.
• Multiplexing dscRNA-seq is desirable for increased scale and reduced technical variability.

Purpose of the Study:

• To introduce demuxlet, a computational tool for sample deconvolution in multiplexed dscRNA-seq.
• To enable accurate assignment of single cells to their original samples within pooled experiments.
• To facilitate the identification of doublets (droplets containing two cells) in pooled samples.

Main Methods:

• Demuxlet leverages natural genetic variation (single-nucleotide polymorphisms or SNPs) to determine sample identity.
• The tool analyzes SNP data from pooled cells to assign singlets and identify doublets.
• Simulations and real-world data were used to validate demuxlet's performance.

Main Results:

• Simulations demonstrated that 50 SNPs per cell are sufficient for high accuracy in pools up to 64 individuals (97% singlet assignment, 92% doublet detection).
• Using real genotyping data, demuxlet achieved >99% singlet recovery and accurate doublet identification in pooled samples.
• The tool was successfully applied to analyze cell-type-specific gene expression changes in lupus patients and perform eQTL analysis.

Conclusions:

• Demuxlet effectively overcomes sample processing limitations and batch effects in dscRNA-seq.
• Multiplexed dscRNA-seq using demuxlet enhances experimental throughput and analytical power.
• The tool facilitates robust differential gene expression and eQTL analyses across multiple individuals.