CiteULike: jyuh's Frishman

MIPS: analysis and annotation of genome information in 2007.

HW Mewes — 2008-05-10T08:57:38-00:00

Nucleic acids research, Vol. 36, No. Database issue. (January 2008)

The Munich Information Center for Protein Sequences (MIPS-GSF, Neuherberg, Germany) combines automatic processing of large amounts of sequences with manual annotation of selected model genomes. Due to the massive growth of the available data, the depth of annotation varies widely between independent databases. Also, the criteria for the transfer of information from known to orthologous sequences are diverse. To cope with the task of global in-depth genome annotation has become unfeasible. Therefore, our efforts are dedicated to three levels of annotation: (i) the curation of selected genomes, in particular from fungal and plant taxa (e.g. CYGD, MNCDB, MatDB), (ii) the comprehensive, consistent, automatic annotation employing exhaustive methods for the computation of sequence similarities and sequence-related attributes as well as the classification of individual sequences (SIMAP, PEDANT and FunCat) and (iii) the compilation of manually curated databases for protein interactions based on scrutinized information from the literature to serve as an accepted set of reliable annotated interaction data (MPACT, MPPI, CORUM). All databases and tools described as well as the detailed descriptions of our projects can be accessed through the MIPS web server (http://mips.gsf.de).

CORUM: the comprehensive resource of mammalian protein complexes.

A Ruepp — 2008-01-18T03:50:41-00:00

Nucleic Acids Res, Vol. 36, No. Database issue. (January 2008)

Protein complexes are key molecular entities that integrate multiple gene products to perform cellular functions. The CORUM (http://mips.gsf.de/genre/proj/corum/index.html) database is a collection of experimentally verified mammalian protein complexes. Information is manually derived by critical reading of the scientific literature from expert annotators. Information about protein complexes includes protein complex names, subunits, literature references as well as the function of the complexes. For functional annotation, we use the FunCat catalogue that enables to organize the protein complex space into biologically meaningful subsets. The database contains more than 1750 protein complexes that are built from 2400 different genes, thus representing 12% of the protein-coding genes in human. A web-based system is available to query, view and download the data. CORUM provides a comprehensive dataset of protein complexes for discoveries in systems biology, analyses of protein networks and protein complex-associated diseases. Comparable to the MIPS reference dataset of protein complexes from yeast, CORUM intends to serve as a reference for mammalian protein complexes.

Protein solubility: sequence based prediction and experimental verification.

Pawel Smialowski — 2007-06-07T15:19:04-00:00

Bioinformatics (6 December 2006)

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.