EvoIF Protein Predictor Fuses Dual Evolutionary Signals Using 0.15% Data
BiologyAIDaily · x · 2026-07-22
The post details EvoIF, a lightweight protein fitness predictor that improves accuracy by explicitly fusing within-family evolutionary profiles with cross-family structural evolutionary signals.
Key Highlights:
- Lightweight & Efficient: Keeps large backbones (ESM-2 & ProteinMPNN) frozen, training only a compact graph encoder. It uses merely 0.15% of the training data required by large baselines while offering faster training.
- Theoretical Innovation: Frames natural evolution as implicit reward maximization, linking masked language modeling (MLM) to maximum-entropy inverse reinforcement learning (IRL) to explain its efficacy in zero-shot prediction.
- Performance: Achieves competitive results on the ProteinGym benchmark (>2.5M mutants), scoring Spearman 0.489 (MSA-free) and 0.519 (MSA-enabled).
Related event: EvoIF Model Enhances Protein Fitness Prediction(2 posts)→
More from Research
- Skyfall GS Uses Flux to Refine Gaussian Splatting, Accepted at ECCV 2026 — ducha_aiki · 2026-09-11
- Could 10k agents discover learning methods beyond backprop, or just tweak existing ones? — SeunghyunSEO7 · 2026-09-11
- Apodex Launches TRACES, First Benchmark for Evaluating 'Discoverative AI' on Real-World Problems — Faheem_uh · 2026-09-11
- TRACES grades the process, not the answer: six-dimension eval for open-ended AI science — Faheem_uh · 2026-09-11
- Apodex launches TRACES, a benchmark grading AI on open-ended discovery instead of known answers — Faheem_uh · 2026-09-11
- Cognition's SWE-2 uses a KKT duality argument in RL to shift the effort Pareto curve — YouJiacheng · 2026-09-11