AUTHOR=Gupta Sudheer , Mittal Parul , Madhu Midhun K. , Sharma Vineet K. TITLE=IL17eScan: A Tool for the Identification of Peptides Inducing IL-17 Response JOURNAL=Frontiers in Immunology VOLUME=8 YEAR=2017 URL=https://www.frontiersin.org/journals/immunology/articles/10.3389/fimmu.2017.01430 DOI=10.3389/fimmu.2017.01430 ISSN=1664-3224 ABSTRACT=
IL-17 cytokines are pro-inflammatory cytokines and are crucial in host defense against various microbes. Induction of these cytokines by microbial antigens has been investigated in the case of ischemic brain injury, gingivitis, candidiasis, autoimmune myocarditis, etc. In this study, we have investigated the ability of amino acid sequence of antigens to induce IL-17 response using machine-learning approaches. A total of 338 IL-17-inducing and 984 IL-17 non-inducing peptides were retrieved from Immune Epitope Database. 80% of the data were randomly selected as training dataset and rest 20% as validation dataset. To predict the IL-17-inducing ability of peptides/protein antigens, different sequence-based machine-learning models were developed. The performance of support vector machine (SVM) and random forest (RF) was compared with different parameters to predict IL-17-inducing epitopes (IIEs). The dipeptide composition-based SVM-model displayed an accuracy of 82.4% with Matthews correlation coefficient = 0.62 at polynomial (