ThatGuyBassam/pyGeno_Scouter

A Streamlit interface that maps clinical symptoms to candidate genes, pathogenic variants, and genomic coordinates using pyGeno, HPO, and Orphanet. Curated for the Moroccan/North African patient population. mostly offline after setup. # 🧬 pyGeno Scouter **Clinical phenotype → genomic coordinates** A Streamlit interface that maps clinical symptoms and disease names to candidate genes, genomic coordinates, transcript isoforms, and known pathogenic variants. Built on top of pyGeno by Tariq Daouda (IRIC Montréal), with a curated database oriented around the North African patient population. --- ## Offline / Online split | Feature | Connection | |---|---| | Phenotype Search (curated, HPO, Orphanet) | ✅ Fully offline | | pyGeno genomic coordinates & sequences | ✅ Fully offline | | Gene Batch — ClinVar pathogenic variants | 🌐 Requires internet | The core use case — typing a symptom and getting candidate genes and genomic data — works with no internet connection after initial setup. ClinVar queries in the Gene Batch tab hit the NCBI E-utilities API live. --- ## What it does You type a clinical phenotype — a symptom, a disease name, a gene name, or a syndrome — and the tool returns: - Candidate genes associated with that phenotype - Genomic coordinates (chromosome, position, strand) from Ensembl GRCh38 - Transcript isoforms with exon-by-exon breakdowns - Protein sequences for each coding isoform - Known pathogenic variants with clinical descriptions A second tab (**Gene Batch + ClinVar**) accepts a list of gene names — one per line or comma-separated — and returns pyGeno coordinates plus live ClinVar pathogenic/likely pathogenic variants for each gene, with a one-click CSV export of all variants. ### Search layers | Layer | Source | Language | Coverage | |---|---|---|---| | Curated | Hand-curated, North Africa-focused | French / English | 13 conditions | | HPO | Human Phenotype Ontology (JAX) | English | 42,553 phenotype terms | | Orphanet | Orphanet rare disease database | French / English | 4,128 diseases | | ClinVar | NCBI ClinVar (live) | English | Pathogenic variants by gene | --- ## The problem it solves pyGeno is a powerful tool for querying personalized genome data programmatically …