Benchmark RadarAMPBench-MT
AMPBench-MT evaluates antimicrobial peptide prediction across binary recognition, species-conditioned potency regression, and endpoint-specific safety readouts (hemolysis, toxicity, selectivity) under a sequence-homology-controlled protocol. It includes 13 source databases and multiple task configurations.
- Released
- 2026-07-28
- Readiness
- Inspectable
- Primary field
- Health & Life Sciences
Why it matters
Existing AMP benchmarks focus on binary recognition, but follow-up decisions need assay-derived evidence. AMPBench-MT provides a joint evaluation with homology-controlled splits to reveal that high binary performance does not guarantee assay-endpoint behavior.
Motivation
Computational AMP discovery is often evaluated through AMP/non-AMP recognition, yet follow-up decisions depend on assay-derived evidence such as target-species potency, hemolysis, toxicity, and selectivity.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.