PMLPR: A novel method for predicting subcellular localization based on recommender systems

Elnaz Mirzaei Mehrabad,Changiz Eslahchi,Reza Hassanzadeh

doi:10.1038/s41598-018-30394-w

Abstract

The importance of protein subcellular localization problem is due to the importance of protein’s functions in different cell parts. Moreover, prediction of subcellular locations helps to identify the potential molecular targets for drugs and has an important role in genome annotation. Most of the existing prediction methods assign only one location for each protein. But, since some proteins move between different subcellular locations, they can have multiple locations. In recent years, some multiple location predictors have been introduced. However, their performances are not accurate enough and there is much room for improvement. In this paper, we introduced a method, PMLPR, to predict locations for a protein. PMLPR predicts a list of locations for each protein based on recommender systems and it can properly overcome the multiple location prediction problem. For evaluating the performance of PMLPR, we considered six datasets RAT, FLY, HUMAN, Du et al., DBMLoc and Höglund. The performance of this algorithm is compared with six state-of-the-art algorithms, YLoc, WOLF-PSORT, prediction channel, MDLoc, Du et al. and MultiLoc2-HighRes. The results indicate that our proposed method is significantly superior on RAT and Fly proteins, and decent on HUMAN proteins. Moreover, on the datasets introduced by Du et al., DBMLoc and Höglund, PMLPR has comparable results. For the case study, we applied the algorithms on 8 proteins which are important in cancer research. The results of comparison with other methods indicate the efficiency of PMLPR.

Highlights

Sub-Cellular Location (SCL) prediction of a protein is a substantial problem in Bioinformatics, because there is a close relationship between the SCL of a protein and its function[1]
Computational methods have their advantages and disadvantages. These methods outperform experimental methods, both in terms of time and cost, but they may not be as accurate as experimental methods. Most of these computational methods focus on the single site SCL of a protein whereas the experimental researches show that many proteins are located in several subcellular locations[11]
We present PMLPR (Protein Multiple Location Prediction based on Recommendation systems) which is a recommendation method based on the bipartite network to predict the SCL of proteins

Summary

Introduction

Sub-Cellular Location (SCL) prediction of a protein is a substantial problem in Bioinformatics, because there is a close relationship between the SCL of a protein and its function[1]. Accurate prediction of subcellular localization helps to identify the potential molecular targets for drugs[2]. Protein data banks are growing rapidly, demanding fast and accurate tools for identifying the SCLs of new proteins. There are two approaches for the protein subcellular localization problem: experimental methods and computational methods. Several experimental approaches such as green fluorescent protein[3], microscopic detection[4] and subcellular proteomics[5] have been already introduced to identify subcellular locations of a protein. Various computational methods have been developed to fill this gap[6,7,8,9,10,11]

Methods

Results

Conclusion