678. USING LARGE-SCALE GENETIC STUDIES TO FIND DRUG TARGETS IN SCHIZOPHRENIA: WHERE ARE WE AT?

Abstract Background One mechanism that holds great promise to drive drug development or drug repurposing in schizophrenia is through leveraging human genetics. Recent retrospective analyses of decades of clinical trial data have demonstrated targets with support from human genetics were more likely to successfully proceed from phase-I trials to launch (Minikel et al., 2024, Nature). These principles also extend to drug repurposing (Reay and Cairns, 2021, Nature Reviews Genetics). However, questions remain how human genetics can be most effectively utilized to stimulate therapeutic development in schizophrenia given its biological complexity and polygenic genetic architecture. Aims & Objectives We aimed to explore the therapeutic tractability of high-confidence schizophrenia risk genes identified by large-scale global genetics consortia. Method We collated 128 high-confidence schizophrenia risk genes prioritized by the Psychiatric Genomics Consortium and the Schizophrenia Exome Sequencing Meta-Analysis Consortium. These high-confidence risk genes were annotated with respect to therapeutic tractability using the Target Central Resource Database (TCRD). We deployed a multimodal pipeline to further refine which of these 128 genes may represent the most promising targets leveraging approaches including multiomics and network biology. Results We found that ~16% of high-confidence schizophrenia risk genes were targets of existing approved drugs or had strong small-molecule binding affinity. Integration with genetically proxied and directly measured gene expression from postmortem brain revealed a subset of these targets which were modulated by existing compounds in a direction predicted to be therapeutically useful. Moreover, ~8% of the high-confidence risk genes lie within protein-protein interaction networks which were found to display an overall enrichment of genetic risk for schizophrenia, suggesting that these targets may be particularly tractable given that evidence related to only a single gene may not be of large enough effect size. In particular, this network approach highlighted that rare and common variant supported L-type voltage-gated calcium channels warrant further consideration. Work is ongoing to further refine which genes may also encode proteins that may be suitable targets for de novo drug discovery. Discussion & Conclusions We conclude that genetic association studies in schizophrenia have yielded signals of pharmacological significance, but more work is needed to fully realise their therapeutic potential.