Prediction of ligand binding sites of proteins is a fundamental and important
task for understanding the function of proteins and screening potential drugs.
Most existing methods require experimentally determined protein holo-structures
as input. However, such structures can be unavailable on novel or less-studied
proteins. To tackle this limitation, we propose LaMPSite, which only takes
protein sequences and ligand molecular graphs as input for ligand binding site
predictions. The protein sequences are used to retrieve residue-level
embeddings and contact maps from the pre-trained ESM-2 protein language model.
The ligand molecular graphs are fed into a graph neural network to compute
atom-level embeddings. Then we compute and update the protein-ligand
interaction embedding based on the protein residue-level embeddings and ligand
atom-level embeddings, and the geometric constraints in the inferred protein
contact map and ligand distance map. A final pooling on protein-ligand
interaction embedding would indicate which residues belong to the binding
sites. Without any 3D coordinate information of proteins, our proposed model
achieves competitive performance compared to baseline methods that require 3D
protein structures when predicting binding sites. Given that less than 50% of
proteins have reliable structure information in the current stage, LaMPSite
will provide new opportunities for drug discovery.

预测蛋白质的配体结合位点对于理解蛋白质功能和筛选潜在药物是一项基础而重要的任务。通过使用蛋白序列和配体分子图作为输入，LaMPSite 模型能够在没有 3D 蛋白质结构信息的情况下预测结合位点，与需要 3D 结构的基准方法相比具有竞争力的性能，这为药物发现提供了新的机会。

蛋白质语言模型驱动的蛋白质序列预测三维配体结合位点

Protein Language Model-Powered 3D Ligand Binding Site Prediction from  Protein Sequence

Comprehensive knowledge of protein-ligand interactions should provide a
useful basis for annotating protein functions, studying protein evolution,
engineering enzymatic activity, and designing drugs. To investigate the
diversity and universality of ligand binding sites in protein structures, we
conducted the all-against-all atomic-level structural comparison of over
180,000 ligand binding sites found in all the known structures in the Protein
Data Bank by using a recently developed database search and alignment
algorithm. By applying a hybrid top-down-bottom-up clustering analysis to the
comparison results, we determined approximately 3000 well-defined structural
motifs of ligand binding sites. Apart from a handful of exceptions, most
structural motifs were found to be confined within single families or
superfamilies, and to be associated with particular ligands. Furthermore, we
analyzed the components of the similarity network and enumerated more than 4000
pairs of ligand binding sites that were shared across different protein folds.

通过数据库搜索和匹配算法，对蛋白数据银行中超过 18 万个配体结合位点进行了全对比，共确定了约 3000 个明确定义的配体结合位点的结构模式，并发现这些结构通常局限于单个家族或超家族，并与特定的配体相关联。同时，分析了相似性网络的组成，并列举了超过 4000 对在不同蛋白质折叠中共享的配体结合位点。