enFoldX tops 8 neoantigen scans and an unseen-peptide benchmark
quaidmorris · x · 2026-07-22
enFoldX wins on one-residue neoantigen discrimination and unseen peptides
The thread adds two more evaluation settings for enFoldX:
- Neoantigen task: when cognate and non-cognate peptides differ by only one residue, enFoldX leads across 8 mutational-scan datasets with a median AUC of 0.71.
- IMMREP25 unseen-peptide challenge: for epitopes at least 4 substitutions away from anything published, enFoldX outperforms every sequence- and structure-based method.
The figures compare enFoldX with sequence and structure baselines and show it consistently on top or near the top.
Related event: enFoldX Turns AlphaFold3 Noise into TCR Recognition Signals(11 posts)→
More from Research
- DeepMind paper: LLMs can't sample randomly on their own — pass random numbers from code instead — rohanpaul_ai · 2026-10-27
- World Model Assigns Weight and Friction to Objects, Then Simulates with a Physics Engine — jonstephens85 · 2026-10-27
- Token Collapse: Study Finds Graph Soft Prompting Tokens in LLMs Fail to Encode Structure — Sapienza · 2026-10-27
- Mathathon 2.0 launches with Terence Tao's SAIR, $1.5M in compute for teams — _sathvikr · 2026-10-27
- The 'sweet lesson' is real: two kinds of ML problems need opposite strategies — mike64_t · 2026-10-27
- PixGR: Pixel-space generative decoding for fast feed-forward 3D reconstruction — kwangmoo_yi · 2026-10-27