The study has been approved by ESID.
Proposed by: C Schuetz, A Klocperk, A Graefe/P Robinson, H Ijspeert
GENOTYPE-PHENOTYPE CORRELATION and PRECISE PHENOTYPIC ANALYSIS UTIILZING MEDICAL ONTOLOGIES IN DEFINED COHORTS – example of the level 2 data from the APDS/SOCS1 subregistries
The Human Phenotype Ontology (HPO) and GA4GH Phenopacket Schema provide a machine-readable framework for encoding clinical phenotypes based on semantic similarity scoring across individuals. This may perspectively allow for longitudinal modelling and outcome prediction using modern machine learning methods. This study will test the feasibility and analytical value using proof-of-principle cohorts (SOCS1 haploinsufficiency and APDS) to:
1) Identify novel genotype-phenotype relationships and other deep phenotypic patterns;
2) Evaluate the potential of SLM/LLM-assisted retrospective data standardization for ontology integration;
3) Establish a methodological framework to be extended for other genetically defined IEI cohorts.
