Protein solubility: sequence based prediction and experimental verification.

@article{citeulike:1370565, abstract = {MOTIVATION: Obtaining soluble proteins in sufficient concentrations is a recurring limiting factor in various experimental studies. Solubility is an individual trait of proteins which, under a given set of experimental conditions, is determined by their amino acid sequence. Accurate theoretical prediction of solubility from sequence is instrumental for setting priorities on targets in large-scale proteomics projects. RESULTS: We present a machine-learning approach called PROSO to assess the chance of a protein to be soluble upon heterologous expression in E. coli based on its amino acid composition. The classification algorithm is organized as a two-layered structure in which the output of primary support vector machine classifiers serves as input for a secondary Naive Bayes classifier. Experimental progress information from the TargetDB database as well as previously published datasets were used as the source of training data. In comparison with previously published methods our classification algorithm possesses improved discriminatory capacity characterized by the Matthews Correlation Coefficient of 0.434 between predicted and known solubility states and the overall prediction accuracy of 72\% (75\% and 68\% for positive and negative class respectively). We also provide experimental verification of our predictions using solubility measurements for 31 mutational variants of two different proteins. AVAILABILITY: A Web server for protein solubility prediction is available at http://webclu.bio.wzw.tum.de:8080/proso. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.}, address = {Department of Genome Oriented Bioinformatics, Technische Universit\"{a}t M\"{u}nchen, Wissenschaftszentrum Weihenstephan, 85350 Freising, Germany.}, author = {Smialowski, Pawel and Martin-Galiano, Antonio J J. and Mikolajka, Aleksandra and Girschick, Tobias and Holak, Tad A A. and Frishman, Dmitrij }, citeulike-article-id = {1370565}, issn = {1460-2059}, journal = {Bioinformatics}, month = {December}, posted-at = {2007-09-12 08:54:19}, priority = {2}, title = {Protein solubility: sequence based prediction and experimental verification.}, url = {http://view.ncbi.nlm.nih.gov/pubmed/17150993}, year = {2006} }

TY - JOUR ID - citeulike:1370565 L3 - citeulike-article-id:1370565 TI - Protein solubility: sequence based prediction and experimental verification. JF - Bioinformatics AD - Department of Genome Oriented Bioinformatics, Technische Universität München, Wissenschaftszentrum Weihenstephan, 85350 Freising, Germany. SN - 1460-2059 N2 - MOTIVATION: Obtaining soluble proteins in sufficient concentrations is a recurring limiting factor in various experimental studies. Solubility is an individual trait of proteins which, under a given set of experimental conditions, is determined by their amino acid sequence. Accurate theoretical prediction of solubility from sequence is instrumental for setting priorities on targets in large-scale proteomics projects. RESULTS: We present a machine-learning approach called PROSO to assess the chance of a protein to be soluble upon heterologous expression in E. coli based on its amino acid composition. The classification algorithm is organized as a two-layered structure in which the output of primary support vector machine classifiers serves as input for a secondary Naive Bayes classifier. Experimental progress information from the TargetDB database as well as previously published datasets were used as the source of training data. In comparison with previously published methods our classification algorithm possesses improved discriminatory capacity characterized by the Matthews Correlation Coefficient of 0.434 between predicted and known solubility states and the overall prediction accuracy of 72% (75% and 68% for positive and negative class respectively). We also provide experimental verification of our predictions using solubility measurements for 31 mutational variants of two different proteins. AVAILABILITY: A Web server for protein solubility prediction is available at http://webclu.bio.wzw.tum.de:8080/proso. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online. AU - Smialowski, Pawel AU - Martin-Galiano, Antonio J AU - Mikolajka, Aleksandra AU - Girschick, Tobias AU - Holak, Tad A AU - Frishman, Dmitrij PY - 2006/12/06/ UR - http://view.ncbi.nlm.nih.gov/pubmed/17150993 ER -