Prediction Of Ligand Binding Residues Research Articles

Protein-ligand binding residue prediction enhancement through hybrid deep heterogeneous learning of sequence and structure data.

Knowledge of protein-ligand binding residues is important for understanding the functions of proteins and their interaction mechanisms. From experimentally solved protein structures, how to accurately identify its potential binding sites of a specific ligand on the protein is still a challenging problem. Compared with structure-alignment-based methods, machine learning algorithms provide an alternative flexible solution which is less dependent on annotated homogeneous protein structures. Several factors are important for an efficient protein-ligand prediction model, e.g. discriminative feature representation and effective learning architecture to deal with both the large-scale and severely imbalanced data. In this study, we propose a novel deep-learning-based method called DELIA for protein-ligand binding residue prediction. In DELIA, a hybrid deep neural network is designed to integrate 1D sequence-based features with 2D structure-based amino acid distance matrices. To overcome the problem of severe data imbalance between the binding and nonbinding residues, strategies of oversampling in mini-batch, random undersampling and stacking ensemble are designed to enhance the model. Experimental results on five benchmark datasets demonstrate the effectiveness of proposed DELIA pipeline. The web server of DELIA is available at www.csbio.sjtu.edu.cn/bioinf/delia/. Supplementary data are available at Bioinformatics online.

A simple iterative method to optimize protein–ligand-binding residue prediction

This article describes an iterative method (IM) for improving protein–ligand-binding residue prediction. Through modifying the binding residue definition in every iteration, this method, step by step, increased the performance of the classifiers used. Using a balanced assessment index (BAI), the classifier optimized by the IM achieved a value of 80.4 that is bigger than the one (66.9) of the initial classifier. According to mean per-instance BAI scores, a direct comparison of methods has been carried out along with an analysis of statistical significance of the differences in performance. The results show that the iterative method (IM) does achieve a higher mean score than the threshold-altering method (TAM) used in our previous study and there is a statistically significant difference between the two methods. The IM has a significant advantage that it is independent of the concrete residue characterization models and learning algorithms, and more extensively applicable. These results indicate that optimizing the binding residue definition is also an effective approach to improve protein–ligand-binding residue prediction.

Identification of binding pockets in protein structures using a knowledge-based potential derived from local structural similarities

BackgroundThe identification of ligand binding sites is a key task in the annotation of proteins with known structure but uncharacterized function. Here we describe a knowledge-based method exploiting the observation that unrelated binding sites share small structural motifs that bind the same chemical fragments irrespective of the nature of the ligand as a whole.ResultsPDBinder compares a query protein against a library of binding and non-binding protein surface regions derived from the PDB. The results of the comparison are used to derive a propensity value for each residue which is correlated with the likelihood that the residue is part of a ligand binding site. The method was applied to two different problems: i) the prediction of ligand binding residues and ii) the identification of which surface cleft harbours the binding site. In both cases PDBinder performed consistently better than existing methods.PDBinder has been trained on a non-redundant set of 1356 high-quality protein-ligand complexes and tested on a set of 239 holo and apo complex pairs. We obtained an MCC of 0.313 on the holo set with a PPV of 0.413 while on the apo set we achieved an MCC of 0.271 and a PPV of 0.372.ConclusionsWe show that PDBinder performs better than existing methods. The good performance on the unbound proteins is extremely important for real-world applications where the location of the binding site is unknown. Moreover, since our approach is orthogonal to those used in other programs, the PDBinder propensity value can be integrated in other algorithms further increasing the final performance.

Homology modeling and function prediction of hABH1, involving in repair of alkylation damaged DNA

Inhibition of DNA repair mechanism through alkylating agents in tumor cells is an important method for cancer treatment. Alkylation damage repair gene AlkB was first reported in E. coli. In human and other mammals eight distinguishing homologs of AlkB were detected and are known as hABH1 to hABH8. Crystal structures of hABH2 and hABH3 elucidated the role of human AlkB homologs involved in DNA and RNA repair pathways. No crystal structure of hABH1 is available for the detailed study on the nature and function of the molecule. In the present work we performed homology modeling and different tertiary structure based study on human AlkB homolog hABH1. hABH1.B99990005.pdb, out of five models generated using the program modeler 9v7 and validated with Ramachandran plot showed that 97.9% residues were in the favored and additional allowed region and less residues in disallowed region, which is the best among all models. Functions of the selected model were studied in terms of cation binding, transition metal ion binding and metal ion binding function with oxidoreductase activity. Two functional sites and one conserved cluster were detected in the protein. Ligand binding residue prediction showed four ligand clusters with 17 ligands in cluster 1. In this cluster seven Fe(2+) heterogen counts were detected. Most significantly, predicted iron-binding motif in hABH1 was found as His231-X-Asp233-XnHis287 which corresponds to His131-XAsp133-Xn-His187 in AlkB of E. coli homologue. This shows the similar pattern of aspartic acid and histidine residues in the functional part of the protein both in human and E. coli. These results can be used further to design inhibitors aiding chemotherapy and cancer related diseases.

Ss-TEA: Entropy based identification of receptor specific ligand binding residues from a multiple sequence alignment of class A GPCRs

BackgroundG-protein coupled receptors (GPCRs) are involved in many different physiological processes and their function can be modulated by small molecules which bind in the transmembrane (TM) domain. Because of their structural and sequence conservation, the TM domains are often used in bioinformatics approaches to first create a multiple sequence alignment (MSA) and subsequently identify ligand binding positions. So far methods have been developed to predict the common ligand binding residue positions for class A GPCRs.ResultsHere we present 1) ss-TEA, a method to identify specific ligand binding residue positions for any receptor, predicated on high quality sequence information. 2) The largest MSA of class A non olfactory GPCRs in the public domain consisting of 13324 sequences covering most of the species homologues of the human set of GPCRs. A set of ligand binding residue positions extracted from literature of 10 different receptors shows that our method has the best ligand binding residue prediction for 9 of these 10 receptors compared to another state-of-the-art method.ConclusionsThe combination of the large multi species alignment and the newly introduced residue selection method ss-TEA can be used to rapidly identify subfamily specific ligand binding residues. This approach can aid the design of site directed mutagenesis experiments, explain receptor function and improve modelling. The method is also available online via GPCRDB at http://www.gpcr.org/7tm/.

FunFOLD: an improved automated method for the prediction of ligand binding residues using 3D models of proteins

BackgroundThe accurate prediction of ligand binding residues from amino acid sequences is important for the automated functional annotation of novel proteins. In the previous two CASP experiments, the most successful methods in the function prediction category were those which used structural superpositions of 3D models and related templates with bound ligands in order to identify putative contacting residues. However, whilst most of this prediction process can be automated, visual inspection and manual adjustments of parameters, such as the distance thresholds used for each target, have often been required to prevent over prediction. Here we describe a novel method FunFOLD, which uses an automatic approach for cluster identification and residue selection. The software provided can easily be integrated into existing fold recognition servers, requiring only a 3D model and list of templates as inputs. A simple web interface is also provided allowing access to non-expert users. The method has been benchmarked against the top servers and manual prediction groups tested at both CASP8 and CASP9.ResultsThe FunFOLD method shows a significant improvement over the best available servers and is shown to be competitive with the top manual prediction groups that were tested at CASP8. The FunFOLD method is also competitive with both the top server and manual methods tested at CASP9. When tested using common subsets of targets, the predictions from FunFOLD are shown to achieve a significantly higher mean Matthews Correlation Coefficient (MCC) scores and Binding-site Distance Test (BDT) scores than all server methods that were tested at CASP8. Testing on the CASP9 set showed no statistically significant separation in performance between FunFOLD and the other top server groups tested.ConclusionsThe FunFOLD software is freely available as both a standalone package and a prediction server, providing competitive ligand binding site residue predictions for expert and non-expert users alike. The software provides a new fully automated approach for structure based function prediction using 3D models of proteins.

Feature-incorporated alignment based ligand-binding residue prediction for carbohydrate-binding modules

Carbohydrate-binding modules (CBMs) share similar secondary and tertiary topology, but their primary sequence identity is low. Computational identification of ligand-binding residues allows biologists to better understand the protein-carbohydrate binding mechanism. In general, functional characterization can be alternatively solved by alignment-based manners. As alignment accuracy based on conventional methods is often sensitive to sequence identity, low sequence identity among query sequences makes it difficult to precisely locate small portions of relevant features. Therefore, we propose a feature-incorporated alignment (FIA) to flexibly align conserved signatures in CBMs. Then, an FIA-based target-template prediction model was further implemented to identify functional ligand-binding residues. Arabidopsis thaliana CBM45 and CBM53 were used to validate the FIA-based prediction model. The predicted ligand-binding residues residing on the surface in the hypothetical structures were verified to be ligand-binding residues. In the absence of 3D structural information, FIA demonstrated significant improvement in the estimation of sequence similarity and identity for a total of 808 sequences from 11 different CBM families as compared with six leading tools by Friedman rank test.

LIBRUS: combined machine learning and homology information for sequence-based ligand-binding residue prediction

Identifying residues that interact with ligands is useful as a first step to understanding protein function and as an aid to designing small molecules that target the protein for interaction. Several studies have shown that sequence features are very informative for this type of prediction, while structure features have also been useful when structure is available. We develop a sequence-based method, called LIBRUS, that combines homology-based transfer and direct prediction using machine learning and compare it to previous sequence-based work and current structure-based methods. Our analysis shows that homology-based transfer is slightly more discriminating than a support vector machine learner using profiles and predicted secondary structure. We combine these two approaches in a method called LIBRUS. On a benchmark of 885 sequence-independent proteins, it achieves an area under the ROC curve (ROC) of 0.83 with 45% precision at 50% recall, a significant improvement over previous sequence-based efforts. On an independent benchmark set, a current method, FINDSITE, based on structure features achieves an ROC of 0.81 with 54% precision at 50% recall, while LIBRUS achieves an ROC of 0.82 with 39% precision at 50% recall at a smaller computational cost. When LIBRUS and FINDSITE predictions are combined, performance is increased beyond either reaching an ROC of 0.86 and 59% precision at 50% recall. Software developed for this study is available at http://bioinfo.cs.umn.edu/supplements/binf2009 along with Supplementary data on the study.

Prediction of protein functional residues from sequence by probability density estimation

The prediction of ligand-binding residues or catalytically active residues of a protein may give important hints that can guide further genetic or biochemical studies. Existing sequence-based prediction methods mostly rank residue positions by evolutionary conservation calculated from a multiple sequence alignment of homologs. A problem hampering more wide-spread application of these methods is the low per-residue precision, which at 20% sensitivity is around 35% for ligand-binding residues and 20% for catalytic residues. We combine information from the conservation at each site, its amino acid distribution, as well as its predicted secondary structure (ss) and relative solvent accessibility (rsa). First, we measure conservation by how much the amino acid distribution at each site differs from the distribution expected for the predicted ss and rsa states. Second, we include the conservation of neighboring residues in a weighted linear score by analytically optimizing the signal-to-noise ratio of the total score. Third, we use conditional probability density estimation to calculate the probability of each site to be functional given its conservation, the observed amino acid distribution, and the predicted ss and rsa states. We have constructed two large data sets, one based on the Catalytic Site Atlas and the other on PDB SITE records, to benchmark methods for predicting functional residues. The new method FRcons predicts ligand-binding and catalytic residues with higher precision than alternative methods over the entire sensitivity range, reaching 50% and 40% precision at 20% sensitivity, respectively. Server: http://frpred.tuebingen.mpg.de. Data sets: ftp://ftp.tuebingen.mpg.de/pub/protevo/FRpred/.