Computational Approaches(1/7)

1. Computational Approaches(1/7) • Computational methods can be divided into four categories: prediction methods based on • (i) The overall protein amino acid composition • (ii) Known targeting sequences • (iii) Sequence homology and/or motifs • (iv) A combination of several sources of information (hybrid methods)

2. Computational Approaches(2/7) (i) The overall protein amino acid composition(1/2) • Nakashima and Nishikawa • Method for discriminating between intracellular and extracellular proteins • Using the distance between the overall amino acid composition vectors • Cedano et al • ProtLock for predicting five classes of subcellular localizations • Extracellular, intracellular, integral membrane, anchored membrane, and nuclear • Reinhardt and Hubbard • NNPSL, an approach using artificial neural networks (ANNs) • Predicting four eukaryotic and three prokaryotic subcellular localizations • Chou et al • SVM-based method for predicting twelve different subcellular localization taking sequence order effects into account

3. Computational Approaches(3/7) (i) The overall protein amino acid composition(2/2) • Huang et al • Using Fuzzy k-NNs algorithm • Describe the dipeptide composition of the whole protein sequence for eleven different localizations • Yu, C.S. (CELLO method) • Prediction of five subcellular localizations in Gram-negative bacteria • Based on the composition of peptides of varying lengths • Andrade et al • First to incorporate structural information into the amino acid composition vectors • Composition of eukaryotic proteins with known structure was used • The rationale behind this approach • The interiors of proteins have stayed fairly constant during evolution

4. Computational Approaches(4/7) (ii) Known targeting sequences • Gunnar von Heijne (TargetP) • The most comprehensive method based on N-terminal targeting sequences • Prediction of chloroplast, mitochondrial, secretory pathway, and other proteins • Claros M.G. (MitoProt and Predotar) • Specifically discriminate chloroplast from mitochondrial proteins • Bannai H. et al (iPsort) • Using knowledge-based rules for prediction based on protein sequence features

5. Computational Approaches(5/7) (iii) Sequence homology and/or motifs • Marcotte et al • Assigns the subcellular localization by constructing phylogenetic profiles of the proteins • Cokol M et al(PredictNLS) • Specialized on recognizing nuclear proteins • Based on a collection of nuclear localization sequences • Lu et al(Proteome Annalyst) • Based on SWISS-PROT keywords and the annotation of homologous proteins

6. Computational Approaches(6/7) (iv) Hybrid methods • Nakai K and Kanehisa(PSORT) • One of the first methods developed for predicting the subcellular localization • Using the overall amino acid composition, N-terminal targeting sequence information, and motifs • This method uses a set of knowledge-based "if-then" rules • Predicts 14 animal and 17 plant subcellular localizations • PSORT II and and PSORT-B-Extensions of the PSORT • Drawid and G • Method that incorporates information about sequence motifs, overall sequence properties and mRNA expression levels • Based on a Bayesian prediction model and was tested on the yeast genome • Guda C et al(MITOPRED) • Specialized for predicting mitochondrial proteins • Based on amino acid composition