A new methods paper out proposing a novel method for improving differential expression analyses in cell-specific RNA-sequencing studies. Jordan Sicherman and colleagues, in a collaboration between groups at UBC, the Campbell Family Mental Health Research Institute, and the Krembil Centre for Neuroinformatics, leverage single-cell sequencing databases to quantify contamination in other RNAseq datasets. Using real-world and simulated data, they implement these measures of contamination to improve downstream bioinformatics analyses and provide a framework for their application.

Click blue book icon above to link to published article.

FIGURE 1 | Illustration for potential off-target cellular contamination in single-cell type RNA-seq studies. (A) A schematic showing cell targeted mRNA sampling in LCM-seq, illustrating targeted sampling of specific cells (green) along with the potential for undesired sampling from surrounding cells (orange). In this case, the desired SST interneuron (green) will be microdissected by cutting around the black path. In this area, processes of the orange cell (orange) lay above and below the targeted cell (green). (B) Expression levels of characteristic cell type specific marker genes for pyramidal cells (Slc17a7) and SST interneurons (Sst) in “Single Cell” (single cell Allen Brain Institute reference data) and two LCM-Seq experiments: “Aging” (GSE119183) and “Stress” (GSE145521). Note the presence of relatively high levels of Slc17a7 in samples from SST, PV, and VIP interneurons in the LCM-Seq Aging and Stress datasets, indicating likely off-target contamination.