Protein subcellular localization prediction

Protein subcellular localization prediction (or just protein localization prediction) involves the computational prediction of where a protein resides in a cell.

In general, it is a computational method to take as input information about a protein, such as a protein sequence, and produce a prediction of a subcellular localization as output.

Prediction of protein subcellular localization is an important component of bioinformatics based prediction of protein function and genome annotation, and it can aid the identification of drug targets.

Background

Most eukaryotic proteins are encoded in the nuclear genome and synthesized in the cytosol, but many need to be further sorted before they reach their final destination. For prokaryotes, proteins are synthesized in the cytoplasm and some must be targeted to other locations such as to a cell membrane or the extracellular environment. Proteins must be localized at their appropriate subcellular compartment to perform their desired function.

Experimentally determining the subcellular localization of a protein is a laborious and time consuming task. Through the development of new approaches in computer science, coupled with an increased dataset of proteins of known localization, computational tools can now provide fast and accurate localization predictions for many organisms. This has resulted in subcellular localization prediction becoming one of the challenges being successfully aided by bioinformatics.

Many protein subcellular localization prediction methods now exceed the accuracy of some high-throughput laboratory methods for the identification of protein subcellular localization.

Particularly, some predictors have been developed that can be used to deal with proteins that may simultaneously exist, or move between, two or more different subcellular locations.

Methods

Several computational tools for predicting the subcellular localization of a protein are publicly available, a few of which are listed below. Predictors can be specialized for proteins in different organisms. Some are specialized for eukaryotic proteins, some for human proteins, and some for plant proteins.

Methods for the prediction of bacterial localization predictors, and their accuracy, have been reviewed.

The development of protein subcellular location prediction has been summarized in two comprehensive review articles.

Tools

Protein localization predicition software

Name	Description	References
APSLAP	Prediction of apoptosis protein sub cellular Localization
Cell-PLoc	A package of web-servers for predicting subcellular localization of proteins in various organisms.
BaCelLo	Prediction of eukaryotic protein subcellular localization. Unlike other methods, the predictions are balanced among different classes and all the localizations that are predicted are considered as equiprobable, to avoid mispredictions.
CELLO	CELLO uses a two-level Support Vector Machine system to assign localizations to both prokaryotic and eukaryotic proteins.
DualPred	Webserver for Predicting plant Proteins Dual-Targeted to Chloroplast and Mitochondria.
ClubSub-P	ClubSub-P is a database of cluster-based subcellular localization (SCL) predictions for Archaea and Gram negative bacteria.
Euk-mPLoc 2.0	Predicting the subcellular localization of eukaryotic proteins with both single and multiple sites.
CoBaltDB	CoBaltDB is a novel powerful platform that provides easy access to the results of multiple localization tools and support for predicting prokaryotic protein localizations.
HSLpred	This method allow to predict subcellular localization of human proteins. This method combines power of composition based SVM models and similarity search techniques PSI-BLAST.
KnowPredsite	A knowledge-based approach to predict the localization site(s) of both single-localized and multi-localized proteins for all eukaryotes.
LOCtree	Prediction based on mimicking the cellular sorting mechanism using a hierarchical implementation of support vector machines. LOCtree is a comprehensive predictor incorporating predictions based on PROSITE/Pfam signatures as well as SwissProt keywords.
LocTree2/3	Subcellular localization prediction for all proteins in all domains of life. LocTree2/3 predicts 3 classes for Archaea, 6 for Bacteria and 18 for Eukaryota
MultiLoc	An SVM-based prediction engine for a wide range of subcellular locations.
PSORT	The first widely used method for protein subcellular localization prediction, developed under the leadership of Kenta Nakai. Now researchers are also encouraged to use other PSORT programs such as WoLF PSORT and PSORTb for making predictions for certain types of organisms (see below). PSORT prediction performances are lower than those of recently developed predictors.
PSORTb	Prediction of bacterial protein localization.
MetaLocGramN	Meta subcellular localization predictor of Gram-negative protein. MetaLocGramN is a gateway to a number of primary prediction methods (various types: signal peptide, beta-barrel, transmembrane helices and subcellular localization predictors). In author's benchmark, MetaLocGramN performed better in comparison to other SCL predictive methods, since the average Matthews correlation coefficient reached 0.806 that enhanced the predictive capability by 12% (compared to PSORTb3). MetaLocGramN can be run via SOAP.
PredictNLS	Prediction of nuclear localization signals.
Proteome Analyst	Prediction of protein localization for both prokaryotes and eukaryotes using a text mining approach.
SCLPred	SCLpred protein subcellular localization prediction by N-to-1 neural networks.
SecretomeP	Prediction of eukaryotic proteins that are secreted via a non-traditional secretory mechanism.
SherLoc	An SVM-based predictor combining MultiLoc with text-based features derived from PubMed abstracts.
SCLAP	An Adaptive Boosting Method for Predicting Subchloroplast Localization of Plant Proteins.
TargetP	Prediction of N-terminal sorting signals.
TMHMM	Prediction of transmembrane helices to identify transmembrane proteins.
WoLF PSORT	An updated version of PSORT/PSORT II for the prediction of eukaryotic sequences.

Application

Determining subcellular localization is important for understanding protein function and is a critical step in genome annotation.

Knowledge of the subcellular localization of a protein can significantly improve target identification during the drug discovery process. For example, secreted proteins and plasma membrane proteins are easily accessible by drug molecules due to their localization in the extracellular space or on the cell surface.

Bacterial cell surface and secreted proteins are also of interest for their potential as vaccine candidates or as diagnostic targets.

Aberrant subcellular localization of proteins has been observed in the cells of several diseases, such as cancer and Alzheimer's disease.

Secreted proteins from some archaea that can survive in unusual environments have industrially important applications.

Databases

Curated protein subcellular locations can be searched in UniProtKB. There are several computationally predicted protein subcellular location databases including the fungal secretome and subcellular proteome knowledgebase (FunSecKB2), the plant secretome and subcellular proteome knowledgebase (PlantSecKB), MetazSecKB for human and animals, and the lactic acid bacterial secretome database. Though there are some inaccuracies in the computational prediction, these databases provide useful resources for further characterizing the protein subcellular locations.