Group Ligands Docking to Protein Pockets
Jiaqi Guan, Jiahan Li, Xiangxin Zhou, Xingang Peng, Sheng Wang, Yunan, Luo, Jian Peng, Jianzhu Ma
TL;DR
This paper introduces extsc{GroupBind}, a novel molecular docking framework that considers multiple ligands simultaneously, leveraging their shared binding tendencies to improve docking accuracy in computational biology.
Contribution
The paper presents a new group-based docking approach with interaction and attention modules, enhancing performance over existing methods.
Findings
Achieved state-of-the-art results on the PDBBind benchmark.
Demonstrated the effectiveness of group ligand consideration in docking.
Improved accuracy over traditional pairwise docking methods.
Abstract
Molecular docking is a key task in computational biology that has attracted increasing interest from the machine learning community. While existing methods have achieved success, they generally treat each protein-ligand pair in isolation. Inspired by the biochemical observation that ligands binding to the same target protein tend to adopt similar poses, we propose \textsc{GroupBind}, a novel molecular docking framework that simultaneously considers multiple ligands docking to a protein. This is achieved by introducing an interaction layer for the group of ligands and a triangle attention module for embedding protein-ligand and group-ligand pairs. By integrating our approach with diffusion-based docking model, we set a new S performance on the PDBBind blind docking benchmark, demonstrating the effectiveness of our proposed molecular docking paradigm.
Peer Reviews
Decision·ICLR 2025 Poster
(1) The writing and organization of this paper are very clear. (2) This paper is intriguing because it is based on the idea of enhancing the binding capability of the current ligand by considering the binding positions of other ligands that target the same protein.
(1) Why are the results of DIFFDOCK in Table 1 worse than those in the original paper, and it seems that the bold text annotations might be inaccurate? (2) From Figure 1, we can see that molecules binding to the same pocket indeed have similar structures, but how many pockets in the dataset exhibit this situation? Is there any statistical data on the number of pockets and the corresponding similar ligands in the PDBBind dataset? (3) During inference, when searching the database for ligands si
1. The idea of leveraging similar binding poses among ligands targeting the same protein is intriguing and biologically relevant. 2. The experimental results suggest that incorporating augmented ligands improves docking performance.
1. The core idea is similar to MCS (Maximum Common Substructure) docking [1, 2], which assume that ligands with similar substructures exhibit similar docking poses. GroupBind, however, assumes all ligands share similar docking poses. Figure 1 depicts highly similar ligands with similar docking structures. Conversely, [3] (Figure 4) illustrates cases where structurally distinct ligands adopt distinct poses. A statistical comparison quantifying the difference between the MCS docking assumption an
The study proposes a new molecu lar docking framework to simultaneously consider multiple ligands docking to a protein.
Overall, this paper offers a certain level of contribution, but the experimental section requires clearer descriptions and further discussion. For the figures in the paper, the authors should provide detailed captions, including explanations of the methods used.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsClick Chemistry and Applications · Computational Drug Discovery Methods · Monoclonal and Polyclonal Antibodies Research
MethodsSoftmax · Attention Is All You Need · ADaptive gradient method with the OPTimal convergence rate · Sparse Evolutionary Training