Author: Yujia Xiang; Quan Zou; Lilin Zhao
Title: VPTMdb: a viral post-translational modification database Document date: 2020_4_2
ID: kl99afiu_7
Snippet: Moreover, PTM was predicted in other species with machine learning approaches (He, et al., 2018; Huang and Li, 2018) . For viral protein serine modification sites identification, we implemented a novel feature-based classifier named VPTMpre into the VPTMdb to provide users with the ability to find viral protein phosphorylated sites. First, we compared several feature extraction methods using support vector machine via a 5-fold cross-validation me.....
Document: Moreover, PTM was predicted in other species with machine learning approaches (He, et al., 2018; Huang and Li, 2018) . For viral protein serine modification sites identification, we implemented a novel feature-based classifier named VPTMpre into the VPTMdb to provide users with the ability to find viral protein phosphorylated sites. First, we compared several feature extraction methods using support vector machine via a 5-fold cross-validation method to obtain the best feature representative strategy. Second, for feature selection, we separately input the features extracted from the previous step into three machine learning predictors (support vector machine, random forest and naïve Bayes). Using the minimum redundancy and maximum relevance (mRMR) algorithm (Hanchuan, et al., 2005) , the predictive performance of these three classifiers on different feature dimensions was compared. Subsequently, the features that performed well in all three classifiers were selected as the most meaningful and significant features and 68-dimensional features were obtained. The results of independent testing showed that 68D features performed author/funder. All rights reserved. No reuse allowed without permission.
Search related documents:
Co phrase search for related documents- classifier perform and feature selection: 1
- classifier perform and machine learning: 1, 2, 3, 4
- classifier perform and support vector: 1, 2
- classifier perform and support vector machine: 1, 2
- classifier perform and vector machine: 1, 2
- classifier predictive performance and machine learning: 1, 2
- classifier predictive performance and machine learning approach: 1, 2
- classifier predictive performance and predictive performance: 1, 2, 3
- classifier predictive performance and support vector: 1
- classifier predictive performance and support vector machine: 1
- classifier predictive performance and vector machine: 1
- cross validation method and feature extraction: 1, 2, 3
- cross validation method and feature extraction method: 1, 2
- cross validation method and feature selection: 1, 2
- cross validation method and independent testing: 1
- cross validation method and machine learning: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11
- cross validation method and machine learning approach: 1
- cross validation method and predictive performance: 1, 2
- cross validation method and support vector: 1, 2, 3, 4, 5, 6, 7
Co phrase search for related documents, hyperlinks ordered by date