QMAP: a benchmark for standardized evaluation of antimicrobial peptide MIC and hemolytic activity regression
Anthony Lavertu, Jacques Corbeil, Pascal Germain · 2026
Read QMAP: a benchmark for standardized evaluation of antimicrobial peptide MIC and hemolytic activity regression by Anthony Lavertu, Jacques Corbeil, Pascal Germain (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13473575.txt.
Description
Anthony Lavertu and two colleagues introduced in 2026 a benchmark for models that predict antimicrobial peptide activity. Antimicrobial peptides offer an alternative to standard antibiotics, but progress has been hard to measure because datasets and test rules differ from study to study. That gap lets models look better than they are. The new benchmark, called QMAP, predicts two properties: potency against microbes (minimum inhibitory concentration), and toxicity to red blood cells (HC50).
QMAP enforces strict homology limits between training and test sets, so performance reflects genuine learning on new peptide families rather than memory of near-duplicates. Re-testing existing potency models under these rules suggested limited progress in six years, weak performance for highly potent peptides, and poor prediction of haemolytic activity. The team released a Python package with a fast Rust engine to ease adoption. For drug discovery, the message is sobering: standardised testing must come first.
Catalog record
| CREATOR | Anthony Lavertu, Jacques Corbeil, Pascal Germain |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 56 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |