Static Recovery Is Not Dynamic Stability: Dynamics-Aware Benchmarking of Protein Motif Scaffolding
cs.LG, q-bio.BM
Submitted: 2026-11-05
Updated: 2026-11-05
Comments: Accepted at NeurIPS 2026 (Evaluations & Datasets Track). 23 pages, 10 figures, including supplementary material. The first three authors contributed equally. BILBY-PRO dataset: https://huggingface.co/datasets/HFMI-PROFOUND/BILBY-PRO ; BILBY-MD code: https://github.com/PROFOUND-HFMI/BILBY-MD
Code: https://github.com/PROFOUND-HFMI/BILBY-MD
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- MotifBench: A standardized protein design benchmark for motif-scaffolding problems
- Fast protein backbone generation with SE(3) flow matching
- Out of Many, One: Designing and Scaffolding Proteins at the Scale of the Structural Universe with Genie 2
- ProteinBench: A Holistic Evaluation of Protein Foundation Models
- Proteina: Scaling Flow-based Protein Structure Generative Models
- DPLM-2: A Multimodal Diffusion Protein Language Model
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis