期刊
BIOLOGY DIRECT
卷 8, 期 -, 页码 -出版社
BIOMED CENTRAL LTD
DOI: 10.1186/1745-6150-8-30
关键词
-
类别
资金
- Council of Scientific and Industrial Research [BSC0121]
- Department of Biotechnology, Govt. of India
- Indian Council of Medical Research (ICMR)
Background: The generation of interferon-gamma (IFN-gamma) by MHC class II activated CD4(+) T helper cells play a substantial contribution in the control of infections such as caused by Mycobacterium tuberculosis. In the past, numerous methods have been developed for predicting MHC class II binders that can activate T-helper cells. Best of author's knowledge, no method has been developed so far that can predict the type of cytokine will be secreted by these MHC Class II binders or T-helper epitopes. In this study, an attempt has been made to predict the IFN-gamma inducing peptides. The main dataset used in this study contains 3705 IFN-gamma inducing and 6728 non-IFN-gamma inducing MHC class II binders. Another dataset called IFNgOnly contains 4483 IFN-gamma inducing epitopes and 2160 epitopes that induce other cytokine except IFN-gamma. In addition we have alternate dataset that contains IFN-gamma inducing and equal number of random peptides. Results: It was observed that the peptide length, positional conservation of residues and amino acid composition affects IFN-gamma inducing capabilities of these peptides. We identified the motifs in IFN-gamma inducing binders/peptides using MERCI software. Our analysis indicates that IFN-gamma inducing and non-inducing peptides can be discriminated using above features. We developed models for predicting IFN-gamma inducing peptides using various approaches like machine learning technique, motifs-based search, and hybrid approach. Our best model based on the hybrid approach achieved maximum prediction accuracy of 82.10% with MCC of 0.62 on main dataset. We also developed hybrid model on IFNgOnly dataset and achieved maximum accuracy of 81.39% with 0.57 MCC. Conclusion: Based on this study, we have developed a webserver for predicting i) IFN-gamma inducing peptides, ii) virtual screening of peptide libraries and iii) identification of IFN-gamma inducing regions in antigen (http://crdd.osdd.net/raghava/ifnepitope/). Reviewers: This article was reviewed by Prof Kurt Blaser, Prof Laurence Eisenlohr and Dr Manabu Sugai.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据