eFindSite: Improved prediction of ligand binding sites in protein models using meta-threading, machine learning and auxiliary ligands

Abstract

Molecular structures and functions of the majority of proteins across different species are yet to be identified. Much needed functional annotation of these gene products often benefits from the knowledge of protein-ligand interactions. Towards this goal, we developed eFindSite, an improved version of FINDSITE, designed to more efficiently identify ligand binding sites and residues using only weakly homologous templates. It employs a collection of effective algorithms, including highly sensitive meta-threading approaches, improved clustering techniques, advanced machine learning methods and reliable confidence estimation systems. Depending on the quality of target protein structures, eFindSite outperforms geometric pocket detection algorithms by 15-40 % in binding site detection and by 5-35 % in binding residue prediction. Moreover, compared to FINDSITE, it identifies 14 % more binding residues in the most difficult cases. When multiple putative binding pockets are identified, the ranking accuracy is 75-78 %, which can be further improved by 3-4 % by including auxiliary information on binding ligands extracted from biomedical literature. As a first across-genome application, we describe structure modeling and binding site prediction for the entire proteome of Escherichia coli. Carefully calibrated confidence estimates strongly indicate that highly reliable ligand binding predictions are made for the majority of gene products, thus eFindSite holds a significant promise for large-scale genome annotation and drug development projects. eFindSite is freely available to the academic community at http://www.brylinski.org/efindsite .

Human Rab8b Protein as a Cancer Target - An In Silico Study

“…Active site (binding pockets) is the region of the protein that is responsible for its activity [29,30]. The active site of Rab8bprotein is predicted using Sitemap of Schrodinger suite, CASTp and efindsite servers [31][32][33][34]. The Grid generated around the active site region, of size 80 × 80 × 80 Å 3 , is used for further virtual screening studies [35].…”

eMatchSite: Sequence Order-Independent Structure Alignments of Ligand Binding Pockets in Protein Models

“…Ligand binding sites were identified in target proteins using e FindSite, a recently developed template-based approach [39], [48]. Similar to structure modeling, binding pocket prediction was performed using only weakly homologous templates with a sequence identity to the target of ≤40%.…”

“…In pocket matching calculations, we used only those proteins, for which the center of each of the best of top five binding sites is predicted within a distance of 8 Å from the experimental pocket center, with the corresponding Matthew's correlation coefficient calculated over binding residues of ≥0.4. The accuracy of ligand binding site prediction certainly depends on the quality of target structures [39], therefore, as shown in Tables 1 and S1, the structural subsets of the SOIPPA, Kahraman, Homogeneous and Steroid datasets (crystal structures, high- and moderate-quality models as well as distorted conformations) comprise different numbers of proteins.…”

“…e FindSite assigns a ligand binding probability to each predicted binding residue in the protein target [39]. The binding probability score, , is a squared difference between the binding probabilities assigned to a pair of residues i and j :where and is a ligand binding probability assigned by e FindSite to i -th residue in the first protein and j -th residue in the second protein, respectively.…”

“…A key feature responsible for its high performance is the extensive use of evolutionary information that can be extracted from only weakly homologous templates complexed with ligands. Essentially, the current work extends ideas already explored in binding pocket prediction by algorithms such as FINDSITE [38] and its successor, e FindSite [39], to address the problem of aligning and quantifying the similarities between ligand binding sites in proteins. The performance of e MatchSite is evaluated using several datasets and compared to other algorithms for binding site matching in large-scale benchmarking calculations.…”