Our work is based on the assumption that a tissue or cell type is important for a particular disease if and only if SNPs near genes with high specific expression in that tissue/cell type are enriched for heritability. This assumption leads to several limitations of our approach. First, when analyzing gene expression data from different tissues, cell type composition can confound the analysis, as we demonstrated in our comparison of brain regions; this makes enrichments of organs such as the esophagus or uterus hard to interpret. Second, tissues/cell types with similar gene expression profiles to a causal tissue/cell type will be identified as relevant to disease, just as SNPs in LD with a causal SNP will be identified as associated to disease in a GWAS; thus, significant tissues/cell types should be cautiously interpreted as the “best proxy” for the truly causal tissue/cell type, which may be unobserved. Third, our focus on nearby SNPs prevents us from leveraging signal from regulatory SNPs that act at longer distances. Our approach is also fundamentally limited by the availability of gene expression data and
