Protein Ubiquitylation And Sumoylation Site Prediction Based On Ensemble And Transfer Learning

Ubiquitylation,a typical post-translational modification(PTM),plays an important role in signal transduction,apoptosis and cell proliferation.A ubiquitylation like PTM,sumoylation also may affect gene mapping,expression and genomic replication.Over the past two decades,machine learning has been widely recognized as an effective computational method for predicting ubiquitylation and sumoylation sites.These existing tools require feature engineering,but failed to provide general interpretable features and probably underutilized the growing amount of data.This prompted us to propose a deep learning-based model that integrates multiple convolution and stacked fully-connected layers of seven supervised learning submodels to extract deep representations from protein sequences and physico-chemical properties.We divided 402 physico-chemical properties into 6 clusters and customized deep networks accordingly for handling the high correlations among one cluster.A stacking ensemble strategy was employed to integrate these deep representations to make prediction.Furthermore,with the advantage of transfer learning,our deep learning model can work well on protein sumoylation site prediction as well after finetuning.Finally,on the high-quality annotated database Uniprot / Swiss-Prot,our model outperformed several well-known ubiquitylation and sumoylation site prediction tools which have been popular used.The precision,sensitivity,specificity,MCC and F1 reached 19.24%,74.49%,74.93%,0.287,0.303 and 28.61%,84.36%,83.99%,0.431,0.426,respectively.Our source code to build the model and all data used to train and test the model are freely available from this website https://github.com/ruiwcoding/Deep Ubi Sumo Pre.