The Golgi apparatus plays a key role in sorting, modifying, packaging and
distributing proteins. Thus, correct identification of the type of a Golgi-resident protein
will provide in-depth insights into its function involved in various biological processes.
In this article, we will briefly outline the prediction of protein subGolgi localizations
using bioinformatics method. We introduced the available databases for training
predicted models. Moreover, three kinds of methods were presented for discriminating
between cis-Golgi and trans-Golgi proteins.
Keywords: Cis-Golgi proteins, trans-Golgi proteins, benchmark dataset,
information parameter, feature selection, increment of diversity, analysis of
variance, binomial distribution, support vector machine, cross-validation.