Reproduce “DynaPPI: A Large-scale Dynamic Protein Dataset for AI-driven Advances in Protein Interactomics” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Jiabao Wei, Zilong Geng, Yuze Wang, Jianjun Li, Ning Ding, Bowen Zhou, Bing Zhang, Zhiyuan Ma
A new large-scale dynamic protein dataset, DynaPPI, is introduced to capture molecular dynamics trajectories of protein complex formation, from dissociated chains to bound states. This resource aims to provide AI models with explicit time-resolved data to learn binding mechanisms, facilitating more accurate prediction of unknown multi-chain protein structures.
Key insights
- Learning dynamic binding pathways, including intermediate states, is essential for AI models to accurately predict multi-chain protein complex structures.
- Mitigating conformational biases by starting simulations from realistic unbound (apo) protein chain configurations is crucial for accurate binding trajectory generation.
- A comprehensive pipeline integrating advanced MD simulations with enhanced sampling techniques can produce high-resolution, multi-modal data on dynamic protein complex formation at scale.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.