Our single-cell TWAS framework was published at the AJHG

Our paper at the AJHG (American Society of Human Genetics) links disease GWASs with single-cell gene expression to uncover cell-specific drivers of atherosclerosis 📢

I often post how studying the effects of genetic variation on gene expression can reveal disease mechanisms & novel drug targets 🧬→💊.

This is a common transcriptome-wide association study (TWAS) framework. However, while previous TWASs focused on the effects of genetic variants on bulk tissue RNA levels, gene expression is regulated at the cell level. Here, we used variants associated with single-cell expression in scRNAseq (sc-eQTLs) data from two independent datasets and performed cis-Mendelian randomization (MR) and colocalization with disease outcomes as a new way of running single-cell TWAS analyses.

Topline findings

1. Despite the much smaller sample size of sc-eQTL datasets for peripheral blood (982 and 120) vs. bulk ones (>30K in eQTLGen), our approach leads to significant discovery gains

   👉 88% of 440 significant cis-MR results would be missed by MR in bulk data

2. We detected 21 sigificant gene/cell-type combinations that were replicated with external sc-eQTL data and also colocalized with signals for coronary artery disease, peripheral artery disease, or atherosclerotic stroke

3. Probably, the most consisitent among them was for LIPA, which met all significance criteria for both coronary artery disease & atherosclerotic stroke.
   👉 This is a signal previously reported in bulk TWAS, which we now more concretely localize to monocytes
   👉 This signal of genetically proxied LIPA expression in monocytes seems to be consistent and relatively specific for atherosclerosis outcomes, as revealed in PheWAS analyses

4. We present follow-up data that high LIPA expression (both mRNA and protein) is also transferred from blood monocytes to plaque macrophages

---

 Congratulations to Anushree Ray for the first paper of her PhD and thanks to all co-authors!