Predicting subcellular localization via protein motif co-occurrence

Abstract

The prediction of subcellular localization of proteins from their primary sequence is a challenging problem in bioinformatics. We have created a Bayesian network localization predictor called PSLT that is based on the combinatorial presence of InterPro motifs and specific membrane domains in human proteins. This probabilistic framework generates a likelihood of localization to all organelles and allows to predict multicompartmental proteins. When used to predict on nine compartments, PSLT achieves an accuracy of 78% as estimated by using a 10-fold cross-validation test and a coverage of 74%. When used to predict the localization of proteins from other closely related species, it achieves a prediction accuracy and a coverage >80%. We compared the localization predictions of PSLT to those determined through GFP-tagging and microscopy for a group of human proteins. We found two general classes of proteins that are mislocalized by the GFP-tagging strategy but are correctly localized by PSLT. This suggests that PSLT can be used in combination with experimental approaches for localization to identify proteins for which additional experimental validation is required. We used our predictor to annotate all 9793 human proteins from SWISS-PROT release 41.25, 16% of which are predicted by PSLT to be present in more than one compartment.

Figure 1

Visualization of maximal motifs. The hypothetical compartment C illustrated in this figure contains two maximal motif sets: {A,B,C} and {A,D}. Nonempty subsets of these motif sets are themselves considered to be motif sets but are not maximal (such as, e.g., {A,B}). {A} is a nonmaximal motif set that is common to both the {A,B,C} and {A,D} maximal motif sets.

Figure 2

Optimized compartment priors. The compartment priors represent the estimate of the percentage of distinct proteins in each compartment. These compartment priors were optimized for PSLT as explained in the Methods section.