Rebelling for a reason: protein structural "outliers"

PASS2 version 6: a database of structure-based sequence alignments of protein domain superfamilies in accordance with SCOPe

The number of protein structures is increasing due to the individual initiatives and rapid development of structure determination techniques. Structure-based sequence alignments of distantly related proteins enable the investigation of structural, evolutionary and functional relationships between proteins and their domains leading to their common evolutionary origin. Protein Alignments organized as Structural Superfamilies (PASS2) is a database that provides such alignments of members of protein domain superfamilies of known structure and with less than 40% sequence identity. PASS2 has been continuously updated in accordance to Structural Classification of Proteins (SCOP), and now Structural Classification of Proteins - extended (SCOPe). The current update directly corresponds to SCOPe 2.06, dealing with 2006 domain superfamilies of known structure and about 14 000 domains. Alignments have been augmented by features such as hidden Markov models, highly conserved residues, structural motifs and gene ontology terms, which are available for download. In this update, we introduce the concepts of 'extreme structural outliers' and 'split superfamilies' as well.

PASS2 database for the structure-based sequence alignment of distantly related SCOP domain superfamilies: update to version 5 and added features

Structure-based sequence alignment is an essential step in assessing and analysing the relationship of distantly related proteins. PASS2 is a database that records such alignments for protein domain superfamilies and has been constantly updated periodically. This update of the PASS2 version, named as PASS2.5, directly corresponds to the SCOPe 2.04 release. All SCOPe structural domains that share less than 40% sequence identity, as defined by the ASTRAL compendium of protein structures, are included. The current version includes 1977 superfamilies and has been assembled utilizing the structure-based sequence alignment protocol. Such an alignment is obtained initially through MATT, followed by a refinement through the COMPARER program. The JOY program has been used for structural annotations of such alignments. In this update, we have automated the protocol and focused on inclusion of new features such as mapping of GO terms, absolutely conserved residues among the domains in a superfamily and inclusion of PDBs, that are absent in SCOPe 2.04, using the HMM profiles from the alignments of the superfamily members and are provided as a separate list. We have also implemented a more user-friendly manner of data presentation and options for downloading more features. PASS2.5 version is available at http://caps.ncbs.res.in/pass2/.

The value of protein structure classification information-Surveying the scientific literature

The Structural Classification of Proteins (SCOP) and Class, Architecture, Topology, Homology (CATH) databases have been valuable resources for protein structure classification for over 20 years. Development of SCOP (version 1) concluded in June 2009 with SCOP 1.75. The SCOPe (SCOP-extended) database offers continued development of the classic SCOP hierarchy, adding over 33,000 structures. We have attempted to assess the impact of these two decade old resources and guide future development. To this end, we surveyed recent articles to learn how structure classification data are used. Of 571 articles published in 2012-2013 that cite SCOP, 439 actually use data from the resource. We found that the type of use was fairly evenly distributed among four top categories: A) study protein structure or evolution (27% of articles), B) train and/or benchmark algorithms (28% of articles), C) augment non-SCOP datasets with SCOP classification (21% of articles), and D) examine the classification of one protein/a small set of proteins (22% of articles). Most articles described computational research, although 11% described purely experimental research, and a further 9% included both. We examined how CATH and SCOP were used in 158 articles that cited both databases: while some studies used only one dataset, the majority used data from both resources. Protein structure classification remains highly relevant for a diverse range of problems and settings.

Automatic classification of protein structures using physicochemical parameters

Protein classification is the first step to functional annotation; SCOP and Pfam databases are currently the most relevant protein classification schemes. However, the disproportion in the number of three dimensional (3D) protein structures generated versus their classification into relevant superfamilies/families emphasizes the need for automated classification schemes. Predicting function of novel proteins based on sequence information alone has proven to be a major challenge. The present study focuses on the use of physicochemical parameters in conjunction with machine learning algorithms (Naive Bayes, Decision Trees, Random Forest and Support Vector Machines) to classify proteins into their respective SCOP superfamily/Pfam family, using sequence derived information. Spectrophores™, a 1D descriptor of the 3D molecular field surrounding a structure was used as a benchmark to compare the performance of the physicochemical parameters. The machine learning algorithms were modified to select features based on information gain for each SCOP superfamily/Pfam family. The effect of combining physicochemical parameters and spectrophores on classification accuracy (CA) was studied. Machine learning algorithms trained with the physicochemical parameters consistently classified SCOP superfamilies and Pfam families with a classification accuracy above 90%, while spectrophores performed with a CA of around 85%. Feature selection improved classification accuracy for both physicochemical parameters and spectrophores based machine learning algorithms. Combining both attributes resulted in a marginal loss of performance. Physicochemical parameters were able to classify proteins from both schemes with classification accuracy ranging from 90-96%. These results suggest the usefulness of this method in classifying proteins from amino acid sequences.