|
| 引用本文: | 刀福英,陈欣欣,林昊.基于物化性质对嗜热蛋白的预测[J].生物信息学,2017,15(1):1-6. |
| DAO Fuying,CHEN Xinxin,LIN Hao.Prediction of thermophilic proteins based on physicochemical properties[J].Chinese Journal of Bioinformatics,2017,15(1):1-6. |
|
| |
|
|
| 本文已被：浏览 2789次 下载 1879次 |
码上扫一扫！ |
|
|
| 基于物化性质对嗜热蛋白的预测 |
|
刀福英,陈欣欣,林昊
|
|
(神经信息教育部重点实验室信息生物学中心(电子科技大学生命科学与技术学院),成都610054)
|
|
| 摘要: |
| 嗜热蛋白在高温下能保持稳定性和活性,是研究蛋白质热稳定性的理想模型,开发一个蛋白质热稳定性识别的方法将对蛋白质工程和蛋白质的设计很有帮助。目前的研究中,氨基酸的组成及其物化性质一直被认为和蛋白质的热稳定性相关。本研究筛选出可靠的数据集,包括915个嗜热蛋白和793个非嗜热蛋白。利用蛋白质氨基酸的物化性质和氨基酸的组成表征嗜热蛋白,将二肽氨基酸组成整合到9组氨基酸物化性质中使蛋白序列公式化。支持向量机5折叠交叉验证表明:当gap=0时,290个特征产生的精度最高,为92.74%。因此说明对于分析蛋白质的热稳定性,所建立的预测模型将是一个很有效的工具。 |
| 关键词: 嗜热蛋白 热稳定性 伪氨基酸组分 氨基酸物化性质 |
| DOI：10.3969/j.issn.1672-5565.2017.01.201606001 |
| 分类号:Q51 |
| 文献标识码:A |
| 基金项目:四川省应用基础研究项目(2015JY0100);中央高校基本业务费(ZYGX2015J144,ZYGX2015Z006)。 |
|
| Prediction of thermophilic proteins based on physicochemical properties |
|
DAO Fuying,CHEN Xinxin,LIN Hao
|
|
(Key Laboratory for Neuro-Information of Ministry of Education, Center for Informational Biology, School of Life Science and Technology,University of Electronic Science and Technology of China,Chengdu 610054,China)
|
| Abstract: |
| Thermophilic proteins can keep stability and activity at high temperature, which are ideal materials to study stability of proteins. Developing a valuable method to identify thermostability of protein would be helpful for protein engineering. In the present study, amino acid composition and physicochemical properties of protein have been thought of being related to the thermostability of protein. A reliable benchmark dataset including 915 thermophilic proteins and 793 non-thermophilic proteins is constructed for training and testing the proposed model in this article. We define protein samples using physicochemical properties and component of amino acid, so we design a descriptor which will combine dipeptide composition with nine physiochemical properties of amino acids. The results by support vector machine (SVM) with 5-fold cross-validation show that the best accuracy is 92.74% by using 290 features when the parameter gap is 0, indicating that our model holds very high potential to become a useful tool for the research on protein thermostability. |
| Key words: Thermophilic proteins Thermostability Pseudo amino acid composition Physico-chemical roperties |
|
|
|
|
