rare-disease-hpo-mapper

What This Skill Does

The rare-disease-hpo-mapper is a specialized OpenClaw AI agent skill designed for clinical phenotype standardization. It bridges the gap between unstructured clinical notes or symptom descriptions and the Human Phenotype Ontology (HPO), which is the international standard for phenotypic abnormalities. By converting raw patient symptoms into codified HPO terms, this skill facilitates accurate rare disease gene diagnosis, clinical trial recruitment, and research consistency. The skill utilizes a structured execution model via scripts/main.py, ensuring that every mapping process is reproducible, audit-ready, and documented with explicit assumptions, making it ideal for precision medicine workflows.

Installation

To integrate this tool into your OpenClaw environment, ensure you have Python 3.10+ installed and the required dependencies defined in the repository. Execute the following command in your terminal:

clawhub install openclaw/skills/skills/aipoch-ai/rare-disease-hpo-mapper

Verify the installation by running the validation command: python -m py_compile scripts/main.py. If successful, the script is ready for use.

Use Cases

1. Clinical Diagnostic Support: Standardize clinical symptoms provided by clinicians into HPO terms to query genomic databases for causative gene mutations.
2. Cohort Phenotyping: Normalize patient symptom lists across multiple research sites to enable large-scale meta-analysis of rare disease variants.
3. Clinical Trial Enrollment: Automate the filtering of patient phenotypes against specific inclusion/exclusion criteria defined by HPO terms, accelerating recruitment.
4. Audit and Compliance: Generate reproducible mapping reports for medical audits where documentation of logic and assumptions is required.

Example Prompts

1. "Map the following clinical summary to HPO terms: Patient presents with persistent macrocephaly, hypotonia, and dysmorphic facial features. Provide the output in JSON format."
2. "Review the symptoms list in patient_data.txt and map them to the HPO library. If a symptom is ambiguous, list it as a low-confidence mapping."
3. "Using the rare-disease-hpo-mapper, generate a standardized phenotype profile for this patient and highlight any symptoms that lack a direct HPO mapping."

Tips & Limitations

• Quality Control: Always review the final HPO mappings. While the tool is highly precise, nuances in medical language can sometimes lead to misinterpretations; use the references/ directory for guidance on complex cases.
• Input Sanitization: Ensure clinical inputs are stripped of identifiable health information (PHI) before processing if working in a non-secure environment, as the script processes text locally.
• Fallback Paths: If the script identifies multiple potential HPO terms for a single input, prioritize manual verification of the candidate terms. Always clarify domain-specific constraints like anatomical specificity or developmental stage before starting the mapping job.