Welcome to the Text to HPO Pipeline project! This tool is designed to map free text descriptions of phenotypic abnormalities to Human Phenotype Ontology (HPO) terms, providing a standardized way to capture and analyze phenotypic data.
⚠️ Warning: This project is currently under development. The repository will be fully functional and publicly available soon. Please check back later for updates.
The text to HPO Pipeline is a Python-based tool that enables the conversion of unstructured clinical descriptions into structured, standardized HPO terms.
Install metamap from https://github.com/AnthonyMRios/pymetamap.
Install pymetamap from https://github.com/AnthonyMRios/pymetamap.
To run main.py, the text_hpo_mapping/config/config.yaml file needs to be adapted:
clinical_data_path contains the path to the input data. A fake example is provided for illustration.
base_dir must contain the base_dir of the metamap install (Note the other metamap settings might need to be adapted as well)
api_key must be set to the user's private UMLS api key. Instruction on where to find the API key can be found here: https://documentation.uts.nlm.nih.gov/rest/authentication.html