A computational phenotyping approach to characterize neurogenetic disorders

PROJECT SUMMARY Over the last decade, there has been an exponential increase in identified genetic causes of neurodevelopmental disorders and epilepsy. With more than 100 genes identified, understanding how phenotypes relate to specific genetic variants is critical, given the clinical complexity of developmental brain disorders. Given that treatment and prognosis is dependent on understanding genotype-phenotype correlations, there is a critical need to better assess clinical features in genetic epilepsies. However, phenotyping is a time-consuming, manual task with limited throughput. To overcome this bottleneck, we have developed a novel approach, based on the Human Phenotype Ontology (HPO), which we have previously applied to SCN2A-related disorders and to STXBP1- related disorders, resulting in knowledge that is already applied clinically. Our long-term goal is to decipher the phenotypic landscape of genetic epilepsies to improve clinical care.Therefore, our objectives are to determine the relationship between genomic variation and epilepsy-related clinical features in a large patient cohort and to identify subgroups within the 20 most common genetic epilepsies that may provide insight into outcomes and treatment responses. We plan to pursue these objectives through two aims. First, we aim to determine the impact of genomic features on epilepsy phenotypes in >9,000 individuals through an HPO-based approach (Aim #1). We will analyze exome data in >13K individuals with trio exome data and >600K HPO terms to assess the relationship between distinct monogenic etiologies and rare variants with clinical epilepsy features, using computational phenotyping tools developed by our team. This will allow for insight into the relationship between genetic etiologies and phenotypic features at a granular scale. Secondly, we aim to define relevant subgroups in genetic epilepsies through phenotype harmonization (Aim #2). We will translate clinical features for the 20 most common genetic epilepsies to HPO terms and perform a semantic similarity analysis to determine whether specific variants have significantly similar clinical features, followed by in-depth chart review. This knowledge will inform the prioritization of variants for functional studies and clinical care. In summary, HPO-based delineation of genetic epilepsies is expected to significantly improve knowledge of genotype-phenotype correlations by adding unmatched detail and power. Our team has previously pioneered computational phenotype analysis in the epilepsies and neurodevelopmental disorders, positioning us uniquely to address these questions. In addition to facilitating research of disease mechanisms by prioritizing variants for work with stem cells or mouse models, for example, our findings will also apply to clinical care by providing an unprecedented level of precision in prognosis and treatment information.