Combining Machine Learning and Homology-Based Approaches to Accurately Predict Subcellular Localization in Arabidopsis

Abstract

A complete map of the Arabidopsis (Arabidopsis thaliana) proteome is clearly a major goal for the plant research community in terms of determining the function and regulation of each encoded protein. Developing genome-wide prediction tools such as for localizing gene products at the subcellular level will substantially advance Arabidopsis gene annotation. To this end, we performed a comprehensive study in Arabidopsis and created an integrative support vector machine-based localization predictor called AtSubP (for Arabidopsis subcellular localization predictor) that is based on the combinatorial presence of diverse protein features, such as its amino acid composition, sequence-order effects, terminal information, Position-Specific Scoring Matrix, and similarity search-based Position-Specific Iterated-Basic Local Alignment Search Tool information. When used to predict seven subcellular compartments through a 5-fold cross-validation test, our hybrid-based best classifier achieved an overall sensitivity of 91% with high-confidence precision and Matthews correlation coefficient values of 90.9% and 0.89, respectively. Benchmarking AtSubP on two independent data sets, one from Swiss-Prot and another containing green fluorescent protein-and mass spectrometry-determined proteins, showed a significant improvement in the prediction accuracy of species-specific AtSubP over some widely used "general" tools such as TargetP, LOCtree, PA-SUB, MultiLoc, WoLF PSORT, Plant-PLoc, and our newly created All-Plant method.

Agrobacterium May Delay Plant Nonhomologous End-Joining DNA Repair via XRCC4 to Favor T-DNA Integration

“…Because XRCC4 is a DSB repair protein, it was predicted to localize to the nucleus (AtSubP) (Kaundal et al, 2010). However, our results surprisingly showed localization in both cytoplasmic and nuclear compartments (see Supplemental Figure 5 online).…”

“…For localization predictions, XRCC4 sequence was processed with AtSubP (Kaundal et al, 2010), a species-specific predictor for protein localization. For localization assay, At-XRCC4 ORF was fused to GFP at the N terminus, and the construct was subsequently agroinfiltrated into N. benthamiana leaves, followed by observation at 48 HAI under a fluorescence microscope.…”

Machine Learning Techniques in Plant Biology

“…A complete map of a plant proteome is clearly a major goal for plant research community in terms of determining the function and regulation of each encoded protein. An integrative support vector machine-based localization predictor called AtSubP was developed (Kaundal et al 2010) which was based on the combinatorial presence of diverse protein features, viz., amino acid composition, sequence-order effects, terminal information, position-specific scoring matrix, and similarity search-based, positionspecific, iterated Basic Local Alignment Search Tool information. The model predicted seven subcellular compartments through fivefold crossvalidation test and achieved an overall sensitivity of 91 % with high-confidence precision and Matthew's correlation coefficient values of 90.9 % and 0.89, respectively.…”

Subcellular Location and Molecular Mobility of Human Cytosolic Sulfotransferase 1C1 in Living Human Embryonic Kidney 293 Cells

“…To date, the machine learning approaches have ranged from use of primary amino acid sequences, localization-specific functional domains, and sorting signals of proteins, to mining of functional annotation data (Mott et al, 2002;Guda and Subramaniam, 2005;Gardy and Brinkman, 2006;Garg et al, 2009). Some of these approaches have been used in combination (Kaundal et al, 2010). In the current studies, machine learning algorithms profiling protein physicochemical properties were used to explore the subcellular location of the native and GFP-tagged SULT1C1 proteins.…”