Mapping Retinal Gene Expression to Cell Classes: Insights into Retinal Circadian Rhythms

Many retinal processes are regulated by circadian rhythms. Previous studies have identified genes with rhythmic retinal expression over a 24 h cycle, but the cell class specificity of many of these genes remains unknown. Here we present a mathematical model, based on existing human retinal single-cell RNA sequencing (scRNA-seq) data and validated using known retinal markers, which assigns cell classes to genes showing uniquely enriched expression in that class. This model assigns retinal cell class specificity to over 10,000 human genes and provides potential novel candidates for retinal cell markers. Additionally, the model enables transcriptomic datasets to be investigated through cell class enrichment analysis. We apply this model to circadian-expressed genes from published mouse ocular transcriptomic data. Specific cell classes were assigned to over 35% of these genes, many of which have little to no previous literature denoting their retinal localisation and role. We also demonstrate significant enrichment of photoreceptor-associated transcripts, many of which are associated with phototransduction and outer segment turnover. The methods and data presented here provide a resource for researchers interested in the cell class specific expression of retinal genes and provide new insight into the cellular basis of retinal circadian rhythms.