BhairPred: prediction of b-hairpins in a protein from multiple alignment information using ANN and SVM techniques

BhairPred: Prediction of β-Hairpins in a Protein from Multiple Alignment Information Using ANN and SVM Techniques Welcome to the official repository for BhairPred, a web server for predicting β-hairpin supersecondary structural motifs in protein sequences using machine learning. This resource is designed to support researchers in protein structure prediction, structural bioinformatics, and computational biology. Web Server: https://webs.iiitd.edu.in/raghava/bhairpred/ Citation Kumar, M., Bhasin, M., Natt, N. K., & Raghava, G. P. S. (2005). BhairPred: prediction of β-hairpins in a protein from multiple alignment information using ANN and SVM techniques. Nucleic Acids Research, 33 (Web Server issue), W154–W159. https://doi.org/10.1093/nar/gki588 About the Tool BhairPred is a web-based prediction server for identifying β-hairpin supersecondary structural motifs in protein sequences. It integrates evolutionary information (PSI-BLAST profiles), predicted secondary structure (PSIPRED), and surface accessibility (NETASA) with two machine-learning classifiers — an Artificial Neural Network (ANN) and a Support Vector Machine (SVM) — to discriminate β-hairpin from non-hairpin β-strand–coil–β-strand (bcb) patterns. The method was trained and tested on: 2,880 non-redundant protein chains from known structures 5,102 β-hairpins and 5,131 non-hairpins of fixed length (17 residues) An additional independent dataset of 534 proteins (Thornton's dataset) from Cruz et al. (2002) 63 CASP6 target proteins for blind evaluation