We compared our findings for Alzheimer’s disease (Table S3, Figure 4B, Figure S14) with a recent study that performed differential expression analysis at the cell type level between 24 Alzheimer’s cases and 24 controls 43 (prefrontal cortex, Brodmann area 10). We tested whether the top 500, top 1000 and top 2000 most differentially expressed genes (no pathology vs pathology) in six different cell types (excitatory neurons, inhibitory neurons, oligodendrocytes, oligodendrocytes precursor cells, astrocyte and microglia) were enriched in genetic associations with Alzheimer’s disease using MAGMA. Consistently with our results, we found that genes differentially expressed in microglia were the most associated with Alzheimer’s disease genetics (Table S11), indicating that our approach appropriately highlight the relevant cell type at a fraction of the cost of a case-control single cell RNA-seq study. As performing case-control single cell RNA-seq studies in the entire nervous system is currently cost prohibitive, the consistency of our results with the case-control study of Alzheimer’s disease suggests that our results could be leveraged to target specific brain regions and cell types in future case-control genomic studies of brain disorders.
