Single-cell RNA-sequencing (SCRS) is a powerful technique to address biological variation by profiling expression in single cells and samples with low RNA input. Previous studies have shown that gene expression is highly correlated across array and standard RNA-sequencing technologies. However, no comparative studies utilizing SCRS and the same starting sample across platforms have been reported.
We compared expression data from tiling arrays and SCRS on RNA harvested from embryonic cholinergic motor neurons (dorsal A), embryonic coelomocytes (macrophage-like cell), and larval dopaminergic neurons from C. elegans. Picogram quantities of total RNA from each sample were amplified using a single-cell protocol to generate double stranded cDNAs, which were then sequenced with the Illumina HiSeq platform. The same RNA samples from each cell type were previously amplified using the NuGEN WT-Ovation Pico protocol and hybridized to Affymetrix tiling arrays.
We compared log2 FPKM counts for each gene with the corresponding RMA-normalized log2 array expression values. These two independent measures of transcript expression are highly correlated (Spearman correlation = 0.75 for coelomocytes, 0.62 for A-class motor neurons, and 0.68 for dopaminergic neurons). Moreover, SCRS data showed several hundred genes that are significantly enriched in each cell type in comparison with existing whole animal RNA-seq from the same developmental stage, and these significantly overlap genes detected as enriched from the tiling array data (p < 5.38e-36 for all three sets).
In sum, the correlation of SCRS to tiling array is high, similar to published comparisons between microarrays and standard RNA-seq where at least one thousand-fold more starting material was used. These results suggest that single-cell RNA-sequencing is a robust tool for gene expression quantification and transcriptome profiling when input material is limiting.
Paul Scheid, Clay Spencer, Michelle Gutwein, Ashish Agarwal, Kristin C. Gunsalus, David Miller III. A Comparison of Single-cell RNA-seq with Gene Expression Microarrays. Advances in Genome Biology and Technology, Marco Island, FL, Feb 2012.