SeqSVM: A Sequence-Based Support Vector Machine Method for Identifying Antioxidant Proteins

Abstract

Antioxidant proteins can be beneficial in disease prevention. More attention has been paid to the functionality of antioxidant proteins. Therefore, identifying antioxidant proteins is important for the study. In our work, we propose a computational method, called SeqSVM, for predicting antioxidant proteins based on their primary sequence features. The features are removed to reduce the redundancy by max relevance max distance method. Finally, the antioxidant proteins are identified by support vector machine (SVM). The experimental results demonstrated that our method performs better than existing methods, with the overall accuracy of 89.46%. Although a proposed computational method can attain an encouraging classification result, the experimental results are verified based on the biochemical approaches, such as wet biochemistry and molecular biology techniques.

Identifying Antioxidant Proteins by Using Amino Acid Composition and Protein-Protein Interactions

“…The experimental results indicated that the groups only containing 3-spaced residue pairs were superior than the others for classification, which also confirmed the conclusion that the 3-gap dipeptides feature in Feng et al (2016) was good for classification. On the other hand, despite previous research showing that CTD could be used to obtain good classification results, such as the combined features of Zhang et al (2016) and Xu et al (2018) with 188D, in fact, the experimental results showed the classification accuracy of the CT groups was higher than that of the CTD groups. Therefore, only containing 3-spaced residue pairs and CT were selected as the methods of feature extraction.…”

“…The dataset we used has been previously used by Feng et al (2016), Xu et al (2018), and Meng et al (2019). We first collected proteins with antioxidant activities from the UniProt database (release 2014_02) according to the following steps: (1) only proteins with experimentally proven antioxidant activities were selected; and (2) ambiguous proteins were excluded, such as those containing non-standard letters like "B, " "X, " and "Z."…”

“…In 2016, an integration method was proposed by Zhang et al (2016), which was applied for predicting antioxidant proteins with mixed features, indicating that protein secondary structure information facilitates the discrimination of target proteins. Then, a method called SeqSVM was presented by Xu et al (2018) employing a 188D feature extraction method. Last year, Meng et al (2019) also utilized a support vector machine with structural features to establish a model to discriminate target proteins.…”

Machine Learning Model for Identifying Antioxidant Proteins Using Features Calculated from Primary Sequences

“…As shown in their studies, AodPred improved the prediction accuracy to 74.79% after the thorough jackknife cross-validation tests. Compared to the accuracy results of previous researches, Xu et al [ 17 ] had presented a computational method called “SeqSVM” in 2018 to predict antioxidant proteins, with the primary sequence features that were observed mainly from the physicochemical properties and sequence information of the protein. The dimensions of the extracted features were 188 with an accuracy of 89.46%.…”

6mA-RicePred: A Method for Identifying DNA N6-Methyladenine Sites in the Rice Genome Based on Feature Fusion

“…SVM is a widely used machine learning algorithm (Ding and Li, 2015;Li et al, 2015;Zeng et al, 2017;Ding et al, 2017a;Zhang et al, 2019;Tan et al, 2019a) and was used in this study to identify 6mA sites in the rice genome. SVM is also widely used in bioinformatics fields (Zou et al, 2016b;Wang et al, 2018;Wei et al, 2018;Xiong et al, 2018;Zeng et al, 2018a;Xu et al, 2018a;Xu et al, 2018b;Xu et al, 2018c;Li et al, 2019). Our experiments showed that SVM was more suitable for the purposes of the present study than were the other algorithms.…”