A Review on Protein Subcellular Localization Prediction using Microscopic Images

Abstract

Subcellular localization of proteins can provide essential information about their functions and structures in cells. With the rapid advancement in modern molecular imaging techniques, bioimages have received considerable attention for automatically assessing the location of proteins in subcellular compartments. Fluorescence microscopy is the most commonly used technique to obtain images of protein patterns in the cell. Knowledge of protein subcellular localization has also proven helpful in early diagnosis of the disease and drug targeting treatment. In this paper, the recent progress in automated Protein Subcellular Localization (PSL) prediction using microscopic images obtained from confocal microscopy has been systematically reviewed. First, an overview of different datasets available for protein subcellular localization prediction has been given. Then, an overview of various machine learning methodologies has been presented, followed by various deep learning techniques applied for detecting protein subcellular localization. Finally, this review summarizes the future prospects and challenges faced in this field.