MK-FSVM-SVDD: A Multiple Kernel-based Fuzzy SVM Model for Predicting DNA-binding Proteins via Support Vector Data Description

Background: Detecting DNA-binding proetins (DBPs) based on biological and chemical methods is time consuming and expensive. Objective: In recent years, the rise of computational biology methods based on Machine Learning (ML) has greatly improved the detection efficiency of DBPs. Method: In this study, Multiple Kernel-based Fuzzy SVM Model with Support Vector Data Description (MK-FSVM-SVDD) is proposed to predict DBPs. Firstly, sex features are extracted from protein sequence. Secondly, multiple kernels are constructed via these sequence feature. Than, multiple kernels are integrated by Centered Kernel Alignment-based Multiple Kernel Learning (CKA-MKL). Next, fuzzy membership scores of training samples are calculated with Support Vector Data Description (SVDD). FSVM is trained and employed to detect new DBPs. Results: Our model is test on several benchmark datasets. Compared with other methods, MK-FSVM-SVDD achieves best Matthew's Correlation Coefficient (MCC) on PDB186 (0.7250) and PDB2272 (0.5476). Conclusion: We can conclude that MK-FSVM-SVDD is more suitable than common SVM, as the classifier for DNA-binding proteins identification.

Download Full-text

Identification of DNA-Binding Proteins by Multiple Kernel Support Vector Machine and Sequence Information

Current Proteomics ◽

10.2174/1570164616666190417100509 ◽

2020 ◽

Vol 17 (4) ◽

pp. 302-310

Author(s):

Yijie Ding ◽

Feng Chen ◽

Xiaoyi Guo ◽

Jijun Tang ◽

Hongjie Wu

Keyword(s):

Support Vector Machine ◽

Dna Binding ◽

Binding Proteins ◽

Dna Binding Proteins ◽

Computational Method ◽

Support Vector ◽

Sequence Information ◽

Data Sets ◽

Multiple Kernel ◽

Kernel Support Vector Machine

Background: The DNA-binding proteins is an important process in multiple biomolecular functions. However, the tradition experimental methods for DNA-binding proteins identification are still time consuming and extremely expensive. Objective: In past several years, various computational methods have been developed to detect DNAbinding proteins. However, most of them do not integrate multiple information. Methods: In this study, we propose a novel computational method to predict DNA-binding proteins by two steps Multiple Kernel Support Vector Machine (MK-SVM) and sequence information. Firstly, we extract several feature and construct multiple kernels. Then, multiple kernels are linear combined by Multiple Kernel Learning (MKL). At last, a final SVM model, constructed by combined kernel, is built to predict DNA-binding proteins. Results: The proposed method is tested on two benchmark data sets. Compared with other existing method, our approach is comparable, even better than other methods on some data sets. Conclusion: We can conclude that MK-SVM is more suitable than common SVM, as the classifier for DNA-binding proteins identification.

Download Full-text

A sequence-based multiple kernel model for identifying DNA-binding proteins

BMC Bioinformatics ◽

10.1186/s12859-020-03875-x ◽

2021 ◽

Vol 22 (S3) ◽

Author(s):

Yuqing Qian ◽

Limin Jiang ◽

Yijie Ding ◽

Jijun Tang ◽

Fei Guo

Keyword(s):

Dna Binding ◽

Binding Proteins ◽

Multiple Kernel Learning ◽

Dna Binding Proteins ◽

Kernel Learning ◽

Detection Methods ◽

Support Vector ◽

Data Sets ◽

System A ◽

Multiple Kernel

Abstract Background DNA-Binding Proteins (DBP) plays a pivotal role in biological system. A mounting number of researchers are studying the mechanism and detection methods. To detect DBP, the tradition experimental method is time-consuming and resource-consuming. In recent years, Machine Learning methods have been used to detect DBP. However, it is difficult to adequately describe the information of proteins in predicting DNA-binding proteins. In this study, we extract six features from protein sequence and use Multiple Kernel Learning-based on Centered Kernel Alignment to integrate these features. The integrated feature is fed into Support Vector Machine to build predictive model and detect new DBP. Results In our work, date sets of PDB1075 and PDB186 are employed to test our method. From the results, our model obtains better results (accuracy) than other existing methods on PDB1075 ($$84.19\%$$ 84.19 % ) and PDB186 ($$83.7\%$$ 83.7 % ), respectively. Conclusion Multiple kernel learning could fuse the complementary information between different features. Compared with existing methods, our method achieves comparable and best results on benchmark data sets.

Download Full-text