MOTIVATION: Predicting protein-ligand binding sites is crucial in studying protein interactions with applications in biotechnology and drug discovery. Two distinct paradigms have emerged for this purpose: sequence-based methods, which leverage protein sequence information, and structure-based methods, which rely on the three-dimensional (3D) structure of the protein. Here, we analyze a hybrid approach that combines the strengths of both paradigms by integrating two recent deep learning architectures: protein language models (pLMs) from the sequence-based paradigm and Graph Neural Networks (GNNs) from the structure-based paradigm. Specifically, we construct a residue-level Graph Attention Network (GAT) model based on the protein's 3D structure that uses pre-trained pLM embeddings as node features. This integration enables us to study the interplay between the sequential information encoded in the protein sequence and the spatial relationships within the protein structure on the model performance. RESULTS: By exploiting a benchmark dataset over a range of ligands and ligand types, we have shown that using the structure information consistently enhances the predictive power of the baselines in absolute terms. Nevertheless, as more complex pLMs are used to represent node features, the relative impact of the structure information represented by the GNN architecture diminishes. The above observations suggest that although the use of the experimental protein structure almost always improves the accuracy of the prediction of the binding site, complex pLMs still contain structural information that leads to good predictive performance even without the use of 3D structure. AVAILABILITY AND IMPLEMENTATION: The datasets generated and/or analyzed during the current study, as well as pretrained models, are available in the following Zenodo link https://zenodo.org/records/15184302. The source code that was used to generate the results of the current study is available in the following GitHub repository https://github.com/hamzagamouh/pt-lm-gnn as well as in the following Zenodo link https://zenodo.org/records/15192327.

MOTIVATION: Structure-based methods for detecting protein-ligand binding sites play a crucial role in various domains, from fundamental research to biomedical applications. However, current prediction methodologies often rely on holo (ligand-bound) protein conformations for training and evaluation, overlooking the significance of the apo (ligand-free) states. This oversight is particularly problematic in the case of cryptic binding sites (CBSs) where holo-based assessment yields unrealistic performance expectations. RESULTS: To advance the development in this domain, we introduce CryptoBench, a benchmark dataset tailored for training and evaluating novel CBS prediction methodologies. CryptoBench is constructed upon a large collection of apo-holo protein pairs, grouped by UniProtID, clustered by sequence identity, and filtered to contain only structures with substantial structural change in the binding site. CryptoBench comprises 1107 structures with predefined cross-validation splits, making it the most extensive CBS dataset to date. To establish a performance baseline, we measured the predictive power of sequence- and structure-based CBS residue prediction methods using the benchmark. We selected PocketMiner as the state-of-the-art representative of the structure-based methods for CBS detection, and P2Rank, a widely-used structure-based method for general binding site prediction that is not specifically tailored for cryptic sites. For sequence-based approaches, we trained a neural network to classify binding residues using protein language model embeddings. Our sequence-based approach outperformed PocketMiner and P2Rank across key metrics, including area under the curve, area under the precision-recall curve, Matthew's correlation coefficient, and F1 scores. These results provide baseline benchmark results for future CBS and potentially also non-CBS prediction endeavors, leveraging CryptoBench as the foundational platform for further advancements in the field. AVAILABILITY AND IMPLEMENTATION: The CryptoBench dataset, including the benchmark model, is available on Open Science Framework-https://osf.io/pz4a9/. The code and tutorial are available at the GitHub repository-https://github.com/skrhakv/CryptoBench/.

Structure-based identification of protein-ligand binding sites plays a crucial role in the initial stages of rational drug discovery pipelines. As machine learning methods are increasingly integrated into the process, a significant challenge arises while training these methods, as labeled data are typically derived from ligand-bound structures. Consequently, these methods struggle to detect binding sites within proteins where the binding site is concealed in the absence of a bound ligand. Here, we explore the possibility of harnessing protein language models to address this issue and compare their performance against state-of-the-art methods, both those specialized in the cryptic binding site (CBS) detection and those that are not. We show that applying pre-trained protein-language models in a relatively straightforward manner enables us to surpass the state-of-the-art of CBS prediction.

SUMMARY: Understanding the mechanism of action of a protein or designing better ligands for it, often requires access to a bound (holo) and an unbound (apo) state of the protein. Resources for the quick and easy retrieval of such conformations are severely limited. Apo-Holo Juxtaposition (AHoJ), is a web application for retrieving apo-holo structure pairs for user-defined ligands. Given a query structure and one or more user-specified ligands, it retrieves all other structures of the same protein that feature the same binding site(s), aligns them, and examines the superimposed binding sites to determine whether each structure is apo or holo, in reference to the query. The resulting superimposed datasets of apo-holo pairs can be visualized and downloaded for further analysis. AHoJ accepts multiple input queries, allowing the creation of customized apo-holo datasets. AVAILABILITY AND IMPLEMENTATION: Freely available for non-commercial use at http://apoholo.cz. Source code available at https://github.com/cusbg/AHoJ-project. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.