Genetic discovery for neuropsychiatric traits in deep phenotype data: novel methods and applications

Summary One of the major problems in human genetics is understanding the genetic causes underlying complex phenotypes, including neuropsychiatric traits such as autism spectrum disorders, bipolar and schizophrenia. Despite tremendous work over the past few decades, it has been frustratingly difficult to get a good understanding of the underlying biological mechanisms in most cases. Nonetheless, large psychiatric genetic studies are beginning to deliver fundamental knowledge about genetic architecture, disease pathways and specific genetic loci for follow-up. Most psychiatric genetic studies to date have focused on individuals of European origin, leading to profound difference in genetic discoveries with limited transferability of results across populations, but also limiting our knowledge about disease pathophysiology in general. Recently, several large projects in neuropsychiatric genetics have focused on collecting and assembling genetic and deep phenotype data in admixed and populations of different geographic origins. Such projects include the Latin American Genomics Consortium (LAGC), the Genomics of Autism in Latino Ancestries (GALA), the Ancestral Population Network (APN), and PsycheMERGE. Most approaches for causal variant discovery fail to account for key complexities that arise in studies of varying geographic origin, including heterogeneity across populations in terms of effect sizes and linkage disequilibrium (LD) structure, and correlations across geographic origins. Furthermore, with meta-analyses with external LD from reference panels being commonly used in genome-wide association studies, certain types of inconsistencies are inevitable. Therefore, existing methods tend to have suboptimal power and can even produce invalid results, i.e., they prioritize non-causal variants. We propose to develop robust fine-mapping tools that model heterogeneity across populations and are robust to inconsistencies in the data. We also propose to leverage a possibly large number of genetically related traits available in electronic health record systems, including diagnoses, lab results and biomarkers with the goal to refine phenotypes and improve power of genetic association studies for psychiatric phenotypes. We further propose to apply these methods to the largest available collections of datasets from various geographic origins for autism, bipolar, schizophrenia and other neuropsychiatric traits, including data from several psychiatric genetics consortia and electronic health record systems. We believe that the proposed research is very timely and leverages modern datasets with the potential to substantially improve our understanding of the biological mechanisms underlying risk to neuropsychiatric diseases, including schizophrenia, autism and related disorders. Project Number: 1R01MH140223-01 | Fiscal Year: 2026 | NIH Institute/Center: National Institute of Mental Health (NIMH) | Principal Investigator: Iuliana Ionita | Institution: COLUMBIA UNIVERSITY HEALTH SCIENCES, NEW YORK, NY | Award Amount: $534,191 | Activity Code: R01 | Study Section: Neurological, Mental and Behavioral Health Study Section[NMBH] View on NIH RePORTER: https://reporter.nih.gov/project-details/11121230