6/6 But the correspondence for most other psychiatric disorders rarely exceeds null expectation suggesting that the genetic mechanisms driving the risk for psychiatric disorders may be distinct from the pathophysiological mechanisms targeted by drugs.

5/6 We show that incorporating information derived from functional bioinformatics data, such as the PPI network can reveal links between GWAS and drugs for some disorders (bipolar disorder and diabetes [used for comparison]).

4/6 We propose a robust and realistic null model for significance testing that expects the identified matches to be specific for the selected group of treatments. Meaning that treatments for a selected disorder are expected to exceed matches for other treatments.

3/6 Here we integrate different functional information from protein-protein interactions, expression quantitative trait loci (eQTL), and spatial gene expression to determine the correspondence between GWAS hits and the genes for current drug targets.

Do psychiatric GWASs find drug targets?

Excited to see our work linking GWAS and pharmacological treatments in psychiatric disorders now published in JAMA Psychiatry jamanetwork.com/journals/jam...
🧬💊🎯

Big thanks to
@bendfulcher.bsky.social
@alexfornito.bsky.social
Mark Bellgrove

🧵👇