Effective prediction of IL-17 inducing peptides using hybrid approach: iIL17pred

BackgroundInterleukin 17 (IL-17) plays a crucial role in regulating the immune system and is associated with numerous diseases. Modulating IL-17 levels has demonstrated potential in mitigating disease symptoms, positioning it as a compelling target for drug development. Therefore, identifying and characterizing novel drug molecules capable of influencing IL-17 levels is critical. Recent advances in therapeutic peptides underscore their promise as attractive drug candidates, inspiring the development of IL-17 modulating peptides. ResultsThis study aimed to enhance existing methods for efficiently classifying IL-17 inducing peptides. Positive and negative datasets were obtained from the Immune Epitope Database, and peptide features were extracted using the pfeature algorithm. A three-stage hybrid approach combining BLAST, MERCI, and machine learning techniques was employed to accurately classify IL-17-inducing peptides. ConclusionsExtensive benchmarking experiments revealed that our proposed method outperforms existing techniques i.e IL17eScan, across key performance metrics, including sensitivity, accuracy, and area under the curve-receiver operating characteristics (AUC-ROC). Our algorithm achieved an accuracy of 88.08% and a Matthews Correlation Coefficient (MCC) of 0.68 on an external dataset, significantly surpassing the accuracy of 78.57% and MCC of 0.57 achieved by existing methods under comparable conditions. This demonstrates a substantial improvement in IL-17 peptide classification. The results are accessible via a user-friendly web server (http://www.soodlab.com/iil17pred/). These findings hold significant potential for predicting IL-17-inducing peptides, which can be further validated experimentally.