Deep learning-based polygenic scores enhance generalizability of psychiatric disorders prediction
Cobuccio, L.; Sigurdsson, A. I.; Georgii Hellberg, K.-L.; Dybdahl Krebs, M.; Meisner, J.; iPSYCH Study Consortium, ; Werge, T.; Benros, M. E.; Schork, A. J.; Rasmussen, S.
Show abstract
Polygenic scores (PGSs) have emerged as promising tools for predicting complex traits from genetic data, however, their predictive performance for psychiatric disorders remains limited and the added value of deep learning (DL) over linear models is underexplored. In this study, we compared our DL model, Genome-Local-Net (GLN), with the linear model bigstatsr in predicting five psychiatric disorders--ADHD, ASD, BIP, MDD, and SCZ--using individual-level genotype data. We further assessed whether combining these internal (individual-based) PGSs with external (GWAS-derived) PGSs and family genetic risk scores (FGRSs) could improve prediction additively or synergistically. While GLN and bigstatsr performed similarly in-sample, GLN showed better generalization on an out-of-sample replication set for ADHD, ASD, and MDD, with an average AUROC gain of 0.026. Integrating internal, external, and family-based scores significantly improved ADHD prediction, though DL-based integration provided no consistent advantage over logistic models. These findings suggest that while DL may enhance generalizability for specific psychiatric traits, linear models remain competitive and effective for genetic risk prediction.
Matching journals
The top 5 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Are psychiatric disorders risk factors for COVID-19 susceptibility and severity? a two-sample, bidirectional, univariable and multivariable Mendelian Randomization study 94%
- Genetic liability to major psychiatric disorders contributes to multi-faceted quality of life outcomes in children and adults 94%
- Clinical autism subscales have common genetic liability that is heritable, pleiotropic, and generalizable to the general population 93%
Similar papers in this journal
Similar papers in this journal
- Integrative multi-omics analysis of genomic, epigenomic, and metabolomics data leads to new insights for Attention-Deficit/Hyperactivity Disorder 94%
- Multi-polygenic scores in psychiatry: from disorder-specific to transdiagnostic perspectives 94%
- TWAS pathway method greatly enhances the number of leads for uncovering the molecular underpinnings of psychiatric disorders 94%
Similar papers in this journal
- Genetic Analysis of Psychosis Biotypes: Shared Ancestry-Adjusted Polygenic Risk and Unique Genomic Associations 94%
- Genetic neurodevelopmental clustering and dyslexia 94%
- Immunological Drivers and Potential Novel Drug Targets for Major Psychiatric, Neurodevelopmental, and Neurodegenerative Conditions 93%
Similar papers in this journal
- Polygenic profiles define aspects of clinical heterogeneity in ADHD 95%
- Identification of shared and differentiating genetic risk for autism spectrum disorder, attention deficit hyperactivity disorder and case subgroups 95%
- Differences in the genetic architecture of common and rare variants in childhood, persistent and late-diagnosed attention deficit hyperactivity disorder 95%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.