Phenotypic match models

Proprietary phenotypic match models assess how closely a gene’s associated disease phenotype matches the patient’s phenotype profile. A degree of similarity is resembled by the phenotypic match score. The phenotypic match score indicates the degree of similarity between the patient’s phenotype set and the phenotype set of the best-matching gene-related disease.

If a gene is linked to multiple diseases, the model compares the patient’s phenotypes with each disease profile individually. The highest score determines the best match, and this disease is automatically displayed in the Evidence graph. The Phenomatch and Phenomeld scores shown in the variant table correspond to this best-matching disease, even if the associated disease is manually updated.

Note: Case reanalysis causes Phenomatch/Phenomeld scores to be recalculated.

Phenotypic match model	Score interpretation
Phenomeld represents the cutting edge of phenotypic matching models from Emedgene. As an advanced iteration of Phenomatch – Powered by AI, it delivers superior gene prioritization.	The score ranges from 0 to 2: A score of 0 indicates no phenotypic match, A score above 0.15 suggests a moderate match, A score above 0.7 indicates a strong phenotypic match.
Phenomatch – Powered by AI identifies genes with disease phenotypes that loosely match the patient’s HPO terms.	The score ranges between 0 and 1 where 1 is the strongest match.
Phenomatch – High Specificity identifies genes with disease phenotypes that exactly match the patient’s HPO terms.	The score ranges between 0 and 1 where 1 is the strongest match.
Phenomatch with Unknown identifies genes of unknown significance based on indirect links to patient phenotypes. These links may include mouse models, gene families, pathways, and other functional associations.	The score ranges between 0 and 1 where 1 is the strongest match.