Machine Learning Identifies Antifungal Resistance Signatures
Harrison et al. (published March 17, 2026) used machine learning on genomic, ecological, and phenotypic data from nearly all Saccharomycotina species to study resistance to eight antifungal drugs. Random forest models predicted resistant yeasts with 54–75% accuracy (fluconazole 75.2%), Erg11 sequence predictors matched genome-wide annotation models, and in silico mutational scans showed clinical resistance variants are destabilizing while common natural variants are more neutral.
Key Points
- 1Predicted resistance: Random forest models classify antifungal resistance with 54–75% accuracy across eight drugs
- 2Identified Erg11: Top predictive residues differ from clinical variants and show lower conservation
- 3Recommend practitioners include diverse natural yeast species to discover broader resistance mechanisms and markers
Scoring Rationale
Strong cross-lineage ML and peer-reviewed evidence; moderate predictive accuracy (54–75%) limits immediate clinical utility, despite insights.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
