New polygenic scoring method separates schizophrenia, bipolar disorder, and depression in genetic analysis
A team has developed differential diagnosis polygenic risk score, a technique that assigns probabilities to multiple psychiatric diagnoses at once instead of testing one disorder against healthy controls. The method models the variance-covariance structure of disorder liabilities and polygenic risk scores. Published in Nature Genetics, the work used Psychiatric Genomics Consortium summary data from three case-control genome-wide association studies covering 41,917 to 173,140 cases per disorder, with a combined training set of 1,048,683 individuals. Testing on held-out cohorts of 11,460 people with equal representation of schizophrenia, bipolar disorder, major depressive disorder, and controls, the model calculated calibrated posterior probabilities for each category. True diagnosis probabilities in the top deciles of predicted probabilities were considerably larger than prior baseline probabilities.
The technique addresses an unmet need in polygenic prediction by distinguishing clinically overlapping psychiatric disorders from one another, not only from controls, and showed appreciable potential for clinical utility in certain settings.
Written by the Genomes desk from the primary source linked above and checked against it. Research use only; not medical advice. Corrections: [email protected].