Comprehensive Prediction and Interpretation of Viral Protein Subcellular Localization

Determining the subcellular localization of viral proteins is indispensable for understanding the activity of the virus and inferring viral protein functions. Although previous studies about predicting viral protein subcellular localization have been developed, they often have the following disadvantages: (i) only focusing on a part of proteins of a species (ii) not considering the presence of multi-location proteins and (iii) lacking interpretability for the results. To address these problems, this paper is firstly predicting all the subcellular localization of the whole viral proteome in the UniProtKB and is interpretable for the results. This paper gives high prediction accuracy for the single-location and multi-location viral proteins by the FUEL-mLoc predictor. More importantly, we did deeply analysis and interpretation of the subcellular localization of all viral proteins. Finally, we have found some essential GO terms which are interpretable for the results and are significant in predicting the subcellular localization of the viral proteins.