Identification of Surface Residues Involved in Protein-Protein Interaction — A Support Vector Machine Approach

Abstract

SummaryWe describe a machine learning approach for sequence-based prediction of protein-protein interaction sites. A support vector machine (SVM) classifier was trained to predict whether or not a surface residue is an interface residue (i.e., is located in the protein-protein interaction surface) based on the identity of the target residue and its 10 sequence neighbors. Separate classifiers were trained on proteins from two categories of complexes, antibody-antigen and proteaseinhibitor. The effectiveness of each classifier was evaluated using leave-one-out (jack-knife) cross-validation. Interface and non-interface residues were classified with relatively high sensitivity (82.3% and 78.5%) and specificity (81.0% and 77.6%) for proteins in the antigen-antibody and protease inhibitor complexes, respectively. The correlation between predicted and actual labels was 0.430 and 0.462, indicating that the method performs substantially better than chance (zero correlation). Combined with recently developed methods for identification of surface residues from sequence information, this offers a promising approach to prediction of residues involved in protein-protein interaction from sequence information alone.

Amino Acid Features for Prediction of Protein-Protein Interface Residues with Support Vector Machines

“…The present approach achieves a substantial improvement of the prediction accuracy from 2.2% to 8.2% on a dataset of 77 individulal proteins collected from the Protein Data Bank compared to the previously reported best prediction accuracies with SVM methods [5,9]. Results of our experiments with the proposed encoding schema confirmed that the accessible surface areas and the evolutionary information of amino acids along the sequences can detect different sequence features to enhance the accuracy of protein-protein interaction residue prediction.…”

“…Chen and Zhou introduced a consensus neural network that combines predictions from multiple models with different levels of accuracy and coverage [4]. Input information derived from single sequences has been used by support vector machine (SVM) for predicting protein-protein interface residues [9]. Recently, Yan et al proposed a two-stage classifier consisting of an SVM and a Bayesian network classifier that identifies interface residues primarily on the basis of sequence information [5].…”

“…Many computational techniques are available in the literature to predict protein-protein interface residues based on different characteristics of known protein-protein interaction sites [4,5,6,7,8,9]. Neural networks use residues in a local neighborhood or a window as inputs to predict protein-protein interface residues at a particular location of an amino acid sequence by finding an appropriate non-linear mapping.…”

Prediction of inter-residue contact clusters from hydrophobic cores

“…To assess our method objectively, two indices, i.e., specificity and sensitivity (Baldi et al, 2000;Yan and Dobbs, 2003;Wang et al, 2006), are introduced in this paper.…”

Generation of Attribute Value Taxonomies from Data for Data-Driven Construction of Accurate and Compact Classifiers

“…• Learning AVT from data for a broad range of real world applications such as census data analysis, text classification, intrusion detection from system log data [13], learning classifiers from relational data [2], and protein function classification [25] and identification of protein-protein interfaces [27].…”