New AI Model ProtLiD^2 Enhances Protein Design with Ligand Conditioning

Researchers have developed ProtLiD$^2$, a novel discrete diffusion model designed for protein sequence-structure co-design that explicitly incorporates ligand information. This model generates both amino acid sequences and discrete structure tokens, leveraging geometry-aware cross-attention to integrate ligand chemical and geometric data. Trained on over a million ligand-protein complexes, ProtLiD$^2$ demonstrates significant improvements in protein design metrics, including global fold confidence and active-site accuracy, outperforming existing methods like Complexa, FAIR, and PocketGen. AI

IMPACT This model advances AI's capability in scientific discovery, potentially accelerating drug development and protein engineering.

RANK_REASON The cluster contains a research paper detailing a new AI model for a scientific domain. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• Complexa
• FAIR
• ProtLiD$^2$

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →