- (Yu et al., 2010) ⇒ Nancy Y. Yu, James R. Wagner, Matthew R. Laird, Gabor Melli, Sébastien Rey, Raymond Lo, Phuong Dao, S. Cenk Sahinalp, Martin Ester, Leonard J. Foster, Fiona S. L. Brinkman. (2010). “PSORTb 3.0: Improved protein subcellular localization prediction with refined localization subcategories and predictive capabilities for all prokaryotes.” In: Bioinformatics, 26(13). doi:10.1093/bioinformatics/btq249
Subject Headings: PSORTb
- It benefited from the PPLRE Project.
PSORTb has remained the most precise bacterial protein subcellular localization (SCL) predictor since it was first made available in 2003. However, the recall needs to be improved and no accurate SCL predictors yet make predictions for Archaea, nor differentiate important localization sub-categories such proteins targeted to a host cell or bacterial organelles. Such improvements should be preferably be encompassed in a freely available web-based predictor that can also be used as a standalone program.
We developed PSORTb v.3.0 with improved recall, higher proteome-scale prediction coverage, and more refined localization sub-categories. It is the first SCL predictor specifically geared for all prokaryotes, including Archaea and also bacteria with atypical membrane/cell wall topologies. It features an improved standalone program, with new batch results delivery system complementing its web interface. We evaluated the most accurate SCL predictors using 5-fold cross validation plus we performed an independent proteomics analysis, showing that PSORTb 3.0 is the most accurate but can benefit from being complemented by Proteome Analyst predictions.
http://www.psort.org/psortb (download open source software or use the web interface).
Computational prediction of bacterial protein subcellular localization (SCL) provides a quick and inexpensive means for gaining insight into protein function, verifying experimental results, annotating newly sequenced bacterial genomes, detecting potential cell surface/secreted drug targets, as well as identifying biomarkers for microbes. In recent years, this area of computational research has achieved an impressive level of precision (Gardy and Brinkman, 2006), allowing SCL prediction tools to be reliably integrated into automated proteome annotation pipelines and to complement analyses of high-throughput proteomics experiments.
PSORTb version 2.0 (Gardy et al., 2005), the most precise bac-terial SCL prediction software (Gardy and Brinkman, 2006), was introduced in 2005, and has been widely used for the SCL predic-tion of individual proteins as well as for whole proteomes. It gene-rates prediction results for five major localizations for Gram-negative bacteria (cytoplasmic, inner membrane, periplasmic, outer membrane, extracellular) and four localizations for Gram-positive bacteria (cytoplasmic, cytoplasmic membrane, cell wall, extracel-lular). Since then, numerous SCL prediction tools have been created for bacteria using a variety of machine learning algorithms: CELLO version 2.0 (Yu et al., 2006) uses multi-layered Support vector machines (SVMs); SLP-Local predicts SCLs based on local composition and distance frequencies of amino acid groups (Mat-suda et al., 2005); PSL101 makes predictions based on amino acid compositions coupled with structural feature conservations (Su et al., 2007), and PSLDoc bases its SVM features on gapped di-peptides (Chang et al., 2008). Other tools such as Gpos-PLoc (Shen and Chou, 2007) and Gneg-PLoc (Chou and Shen, 2006) make predictions for bacterial proteins by clustering Swiss-Prot proteins with annotated SCLs based on their GO terms and amino acid properties using the K-nearest neighbor (KNN) algorithm. Some methods, such as SubcellPredict and HensBC, combine mul-tiple classifying algorithms in order to boost the prediction perfor-mance (Niu et al., 2008; Bulashevska and Eils, 2006). LocateP (Zhou et al., 2008) and Augur (Billion et al., 2006) differentiate between different types of membrane-anchored, cell wall anchored ...,
|2010 PSORTb3||Nancy Y. Yu|
James R. Wagner
Matthew R. Laird
S. Cenk Sahinalp
Leonard J. Foster
Fiona S. L. Brinkman
|PSORTb 3.0: Improved protein subcellular localization prediction with refined localization subcategories and predictive capabilities for all prokaryotes||http://www.cs.sfu.ca/~ester/papers/PSORTb3.final.pdf||10.1093/bioinformatics/btq249||2010|
|Author||Nancy Y. Yu +, James R. Wagner +, Matthew R. Laird +, Gabor Melli +, Sébastien Rey +, Raymond Lo +, Phuong Dao +, S. Cenk Sahinalp +, Martin Ester +, Leonard J. Foster + and Fiona S. L. Brinkman +|
|title||PSORTb 3.0: Improved protein subcellular localization prediction with refined localization subcategories and predictive capabilities for all prokaryotes +|