Characterization and Identification of Lysine Succinylation Sites based on Deep Learning Method

Abstract

Succinylation is a type of protein post-translational modification (PTM), which can play important roles in a variety of cellular processes. Due to an increasing number of site-specific succinylated peptides obtained from high-throughput mass spectrometry (MS), various tools have been developed for computationally identifying succinylated sites on proteins. However, most of these tools predict succinylation sites based on traditional machine learning methods. Hence, this work aimed to carry out the succinylation site prediction based on a deep learning model. The abundance of MS-verified succinylated peptides enabled the investigation of substrate site specificity of succinylation sites through sequence-based attributes, such as position-specific amino acid composition, the composition of k-spaced amino acid pairs (CKSAAP), and position-specific scoring matrix (PSSM). Additionally, the maximal dependence decomposition (MDD) was adopted to detect the substrate signatures of lysine succinylation sites by dividing all succinylated sequences into several groups with conserved substrate motifs. According to the results of ten-fold cross-validation, the deep learning model trained using PSSM and informative CKSAAP attributes can reach the best predictive performance and also perform better than traditional machine-learning methods. Moreover, an independent testing dataset that truly did not exist in the training dataset was used to compare the proposed method with six existing prediction tools. The testing dataset comprised of 218 positive and 2621 negative instances, and the proposed model could yield a promising performance with 84.40% sensitivity, 86.99% specificity, 86.79% accuracy, and an MCC value of 0.489. Finally, the proposed method has been implemented as a web-based prediction tool (CNN-SuccSite), which is now freely accessible at http://csb.cse.yzu.edu.tw/CNN-SuccSite/.

Characterization and identification of lysine crotonylation sites based on machine learning method on both plant and mammalian

“…Finally, each element in the feature matrix PSSM x was normalized using a Sigmoid function 30 , and �(x) can be written as: w = 31 amino acid of each sequence in this study, so the sequence length ranges from −15 to +15 in this study.…”

iDPGK: characterization and identification of lysine phosphoglycerylation sites based on sequence-based features

“…After the sequence extraction process, we focused on the analysis of sequence-based features, and then each sequence fragment was encoded based on the investigated features. The following sequence-based features are widely employed for analysis and prediction of various types of PTM sites in the enormous amount of research [18,20,21]: amino acid composition (AAC), positively charged amino acid composition (PCAAC), amino acid pair composition (AAPC), BLOSUM62 scoring matrix (B62) and position-specific scoring matrix (PSSM). In this study, the phosphoglycerylated sequences should be transformed into numeric vectors based on the above features to construct a supervised learning model.…”

“…There is an 842-dimensional feature vector made up of sequential and statistical features, which was composed by three types of features including AAC, AAPC and PSSM. By referring to the CNN-SuccSite method [20], all the features were sorted and ranked according to F-score on training dataset prior to construction of predictive models. Furthermore, the sequential forward selection (SFS) [28] is a type of stepwise regression which involves beginning with an empty model and testing the addition of each variable, then adding the variables one at a time until none improves the model to a statistically significant extent.…”

iDPGK: Characterization and Identification of Lysine Phosphoglycerylation Sites Based on Sequence-Based Features

“…As in our previous study, the sequence-based features including amino acid composition (AAC), amino acid pair composition (AAPC), BLOSUM62 scoring matrix (B62) and position-specific scoring matrix (PSSM) that were used for the identification of protein carbonylation sites [14]. Note that all these sequence-based features have widely been employed for analysis and prediction of various types of PTM sites in the enormous amount of research [15][16][17]. In this study, the phosphoglycerylated sequences should be transformed into numeric vectors based on the above features to construct a supervised learning model.…”

“…There is an 842-dimensional feature vector made up of sequential and statistical features, which was composed by three types of features including AAC, AAPC and PSSM. By referring to the CNN-SuccSite method [15], all the features were sorted and ranked according to F-score on training dataset prior to construction of predictive models.…”