Machine Learning for Protein Subcellular Localization Prediction

5 From single- to multi-location

Instead of only determining subcellular localization of single-label proteins, this chapter will focus on predicting both single- and multi-location proteins. Biological significance of multi-location proteins will be first elaborated, followed by a brief introduction of existing algorithms for multi-label classification. Subsequently, three multi-label predictors, namely mGOASVM, AD-SVM, and mPLR-Loc, are presented for multi-location protein subcellular localization.

5.1 Significance of multi-location proteins

Previous chapters show that remarkable progress in the development of computational methods has been made in the past decades for protein subcellular localization. Unfortunately, most of the existing methods ...