My go-to resources for
annotating cell identity

Single cell RNA-seq data sets are too large and complex for the human brain to visualize or understand. Dimension reduction approaches find structure within data then reduce these to discrete dimensions. It is standard to visualize single cell data in two dimensions on a UMAP or t-SNE plot.

The largest source of variability within these data is typically driven by cell identity, thus cells can be assigned to clusters that correspond to cell class. Differential expression analysis comparing each cluster generates a list of marker genes that correspond to cell identity. There are several automated approaches for annotating cell clusters, but manual assignment is typically required. Below I share some tools that I have found useful for annotating cell identity: