SVM-SulfoSite: A support vector machine based predictor for sulfenylation sites

Leslie B Poole,Hussam J Al-Barakati,Robert H Newman,Evan W Mcconnell,Dukka B Kc,Leslie M Hicks

doi:10.1038/s41598-018-29126-x

Abstract

Protein S-sulfenylation, which results from oxidation of free thiols on cysteine residues, has recently emerged as an important post-translational modification that regulates the structure and function of proteins involved in a variety of physiological and pathological processes. By altering the size and physiochemical properties of modified cysteine residues, sulfenylation can impact the cellular function of proteins in several different ways. Thus, the ability to rapidly and accurately identify putative sulfenylation sites in proteins will provide important insights into redox-dependent regulation of protein function in a variety of cellular contexts. Though bottom-up proteomic approaches, such as tandem mass spectrometry (MS/MS), provide a wealth of information about global changes in the sulfenylation state of proteins, MS/MS-based experiments are often labor-intensive, costly and technically challenging. Therefore, to complement existing proteomic approaches, researchers have developed a series of computational tools to identify putative sulfenylation sites on proteins. However, existing methods often suffer from low accuracy, specificity, and/or sensitivity. In this study, we developed SVM-SulfoSite, a novel sulfenylation prediction tool that uses support vector machines (SVM) to identify key determinants of sulfenylation among five feature classes: binary code, physiochemical properties, k-space amino acid pairs, amino acid composition and high-quality physiochemical indices. Using 10-fold cross-validation, SVM-SulfoSite achieved 95% sensitivity and 83% specificity, with an overall accuracy of 89% and Matthew’s correlation coefficient (MCC) of 0.79. Likewise, using an independent test set of experimentally identified sulfenylation sites, our method achieved scores of 74%, 62%, 80% and 0.42 for accuracy, sensitivity, specificity and MCC, with an area under the receiver operator characteristic (ROC) curve of 0.81. Moreover, in side-by-side comparisons, SVM-SulfoSite performed as well as or better than existing sulfenylation prediction tools. Together, these results suggest that our method represents a robust and complementary technique for advanced exploration of protein S-sulfenylation.

Highlights

Redox-dependent signalling plays a critical role in physiological processes such as aging and the immune response, as well as in a number of pervasive diseases, including cancer, Alzheimer’s disease, cardiovascular disease and diabetes[1,2,3]
To develop a robust sulfenylation site prediction tool that is able to identify putative sulfenylation sites using only the primary amino acid sequence as input, we first compiled training and independent test sets similar to those described by Xu et al.[18]
We selected five features—binary encoding (BE), 14 types of physicochemical amino acid properties (AAindex), k-spaced amino acid pairs (KSAAP), amino acid composition (AAC) and high-quality indices (HQI)—and generated a unique vector for each feature

Summary

Introduction

Redox-dependent signalling plays a critical role in physiological processes such as aging and the immune response, as well as in a number of pervasive diseases, including cancer, Alzheimer’s disease, cardiovascular disease and diabetes[1,2,3]. Hasan et al developed a predictor named SulCysSite that uses a random forest-based strategy to identify key parameters related to sulfenylation from among four features. Using this approach, the authors observed SN, SP and MCC scores of 62%, 81%, and 0.45, respectively, upon 10-fold cross-validation[23]. SVM-SulfoSite combines multiple features, including physiochemical properties, amino acid composition and high-quality indices, with novel classifier algorithms and an SVM-based machine learning strategy to predict sulfenylation sites in proteins. Based on evaluation using both 10-fold cross-validation and an independent dataset, SVM-SulfoSite compares favourably to existing sulfenylation site prediction tools with regard to accuracy, sensitivity, specificity. SVM-SulfoSite represents a robust, accessible sulfenylation prediction tool that promises to provide additional insights into the regulation and biological consequences of protein S-sulfenylation

Methods

Results

Conclusion