1 A and is described in detail in Materials and Methods. A schematic of the LBias workflow is shown in Fig. Results Ligand-Binding Residue Prediction.Ī “structural BLAST” approach ( 8) is used to predict potential ligand-binding residues. Notably, using a naive Bayesian network to combine LT-scanner with a sequence-based approach yielded further improvement in performance. Its encouraging performance and its ability to account for binding specificity among closely related proteins suggests that the method can be used effectively for both drug repurposing and “off-target” prediction (i.e., unintended targets of a given drug). LT-scanner was used to predict known target human proteins of 200 Food and Drug Administration (FDA)-approved drugs that were extracted from drug–target databases ( 24– 27). Others involve ligand-based quantitative structure–activity relationships ( 18– 22), although a recently developed approach, FINDSITE comb ( 23), combines both template-based and chemical similarity-based approaches. A number of methods use binding site similarities to predict targets ( 16, 17). Several computational approaches have been developed previously for target protein prediction. LT-scanner takes a ligand–protein complex structure as input and scans through a protein structure database to identify proteins that might bind to that ligand ( Fig. With this goal in mind, we developed ligand–target scanner (LT-scanner) a method to predict, on a genome-wide scale, target proteins for a given ligand based on the LBias scoring function. The success of LBias suggests that its representation of specific types of protein–ligand interactions might be effective in the prediction of the proteins that bind to a particular ligand (the ligand’s “targets”). MetaPocket 2.0 ( 6) and COACH ( 7) are “metaservers” that combine results from a range of structure-based approaches using machine learning. FTsite ( 5) uses docking to probe a protein surface with various types of chemical groups and uses an empirical scoring function to identify surface patches that might favorably interact with those groups. However, since there can be concave regions on a protein surface that do not bind small molecules, or conversely, convex/flat regions that do, programs such as ConCavity ( 3) and LIGSITE CSC ( 4) combine pocket finding algorithms with sequence conservation information. One involves the identification of binding pockets on the protein surface based for example on surface curvature ( 1, 2). Existing structure-based methods for binding site prediction fall into distinct categories. Predicted ligand-binding residues can be used to guide in silico screening of chemical libraries using docking or other approaches. Problem i, the prediction of residues on a protein surface that bind ligands, has been widely studied. The approach we describe is extendable to the more general problem of identifying binding partners of known ligands even if they do not appear in a structurally determined complex, although this will require the integration of methods that combine protein structure and chemical compound databases. Combining sequence with structural information further improves LT-scanner performance. The specificity of the scoring function that we use is demonstrated by the ability of LT-scanner to identify the known targets of FDA-approved kinase inhibitors based on templates involving other kinases. LT-scanner’s performance is evaluated based on its ability to identify known targets of Food and Drug Administration (FDA)-approved drugs and it too proves to be highly effective. The algorithm that we present, LBias, is shown to compare very favorably to existing algorithms for binding site residue prediction. The overall approach is first tested based on its ability to predict the residues on the surface of a protein that are likely to bind small-molecule ligands. A scoring function that rapidly accounts for binding site similarities between the template and the proteins being scanned is a crucial feature of the method. We report a template-based method, LT-scanner, which scans the human proteome using protein structural alignment to identify proteins that are likely to bind ligands that are present in experimentally determined complexes.
0 kommentar(er)
