Convergent proteogenomic evidence prioritises five causal proteins and new drug targets for major psychiatric disorders

Distinguishing causal biology from confounding or downstream consequences of psychiatric disorders remains a key barrier for drug development in psychiatry. We performed a large-scale proteogenomic investigation using a sequential triangulation framework integrating plasma proteomics, Mendelian randomisation, genetic colocalisation, transcriptomics, rare-variant analyses, and clinical phenotyping to identify causal proteins and prioritise therapeutic targets for depression, anxiety, bipolar disorder, and psychotic disorders. Using 2,920 plasma proteins measured in 52,615 UK Biobank participants, we identified 830 protein-disorder associations involving 574 proteins. Mendelian randomisation and colocalisation prioritised 26 proteins with putative causal effects, of which 17 are potentially druggable. Integrating multi-omic and phenotypic evidence ultimately resulted in five high-confidence causal candidates: DDR1 and LTB for depression, DDR1 for anxiety, DSG3 and PBXIP1 for bipolar disorder, and PDIA3 for psychosis. These findings provide convergent evidence implicating neuroimmune and neurodevelopmental pathways in psychiatric disorder biology, while also identifying potentially tractable targets for therapeutic development.