Abstract:BackgroundUbiquitination, which is also called “lysine ubiquitination”, occurs when an ubiquitin is attached to lysine (K) residues in targeting proteins. As one of the most important post translational modifications (PTMs), it plays the significant role not only in protein degradation, but also in other cellular functions. Thus, systematic anatomy of the ubiquitination proteome is an appealing and challenging research topic. The existing methods for identifying protein ubiquitination sites can be divided into two kinds: mass spectrometry and computational methods. Mass spectrometry-based experimental methods can discover ubiquitination sites from eukaryotes, but are time-consuming and expensive. Therefore, it is priority to develop computational approaches that can effectively and accurately identify protein ubiquitination sites.ResultsThe existing computational methods usually require feature engineering, which may lead to redundancy and biased representations. While deep learning is able to excavate underlying characteristics from large-scale training data via multiple-layer networks and non-linear mapping operations. In this paper, we proposed a deep architecture within multiple modalities to identify the ubiquitination sites. First, according to prior knowledge and biological knowledge, we encoded protein sequence fragments around candidate ubiquitination sites into three modalities, namely raw protein sequence fragments, physico-chemical properties and sequence profiles, and designed different deep network layers to extract the hidden representations from them. Then, the generative deep representations corresponding to three modalities were merged to build the final model. We performed our algorithm on the available largest scale protein ubiquitination sites database PLMD, and achieved 66.4% specificity, 66.7% sensitivity, 66.43% accuracy, and 0.221 MCC value. A number of comparative experiments also indicated that our multimodal deep architecture outperformed several popular protein ubiquitination site prediction tools.ConclusionThe results of comparative experiments validated the effectiveness of our deep network and also displayed that our method outperformed several popular protein ubiquitination site prediction tools. The source codes of our proposed method are available at https://github.com/jiagenlee/deepUbiquitylation.

GPS-Uber: a hybrid-learning framework for prediction of general and E3-specific lysine ubiquitination sites

Prediction Of Ubiquitination Sites By Using The Composition Of K-Spaced Amino Acid Pairs

DeepUbi: a deep learning framework for prediction of ubiquitination sites in proteins

Hcksaap_ubsite: Improved Prediction of Human Ubiquitination Sites by Exploiting Amino Acid Pattern and Properties

Large-scale prediction of protein ubiquitination sites using a multimodal deep architecture

Prediction of lysine ubiquitination with mRMR feature selection and analysis

Prediction of Lysine Ubiquitylation with Ensemble Classifier and Feature Selection

Incorporating Key Position and Amino Acid Residue Features to Identify General and Species-Specific Ubiquitin Conjugation Sites

Computational Prediction of Ubiquitination Proteins Using Evolutionary Profiles and Functional Domain Annotation

A New Scheme to Characterize and Identify Protein Ubiquitination Sites

iUbiq-Lys: prediction of lysine ubiquitination sites in proteins by extracting sequence evolution information via a gray system model

UbiSitePred: A novel method for improving the accuracy of ubiquitination sites prediction by using LASSO to select the optimal Chou's pseudo components

Computational prediction of ubiquitination protein using evolutionary profiles and functional domains

GPS-PUP: computational prediction of pupylation sites in prokaryotic proteins.

Towards more accurate prediction of ubiquitination sites: a comprehensive review of current methods, tools and features

An Ensemble Deep Learning based Predictor for Simultaneously Identifying Protein Ubiquitylation and SUMOylation Sites

Prediction and Analysis of Multiple Protein Lysine Modified Sites Based on Conditional Wasserstein Generative Adversarial Networks

UPFPSR: a Ubiquitylation Predictor for Plant Through Combining Sequence Information and Random Forest

Prediction of Protein Ubiquitination Sites in Arabidopsis thaliana

Using WPNNA classifier in ubiquitination site prediction based on hybrid features.

EPuL: an Enhanced Positive-Unlabeled Learning Algorithm for the Prediction of Pupylation Sites