A fast SCOP fold classification system using content-based E-Predict algorithm

Abstract

Background: Domain experts manually construct the Structural Classification of Protein (SCOP) database to categorize and compare protein structures. Even though using the SCOP database is believed to be more reliable than classification results from other methods, it is labor intensive. To mimic human classification processes, we develop an automatic SCOP fold classification system to assign possible known SCOP folds and recognize novel folds for newly-discovered proteins.

Dimensionality reduction in computational demarcation of protein tertiary structures

“…It is notable that though we have reduced the dimension of quantitative feature-vector representation of protein tertiary structures to at the most seven, the accuracy of structural classification we get is comparable to or better than that of Chi et al [12,13]. In the case of common fold types representing a structural class, the dimension as less than five is adequate for predictive classification with high accuracy.…”

“…Considering that SCOP database does finer structural classifications at different fold levels and is also the basis/ yardstick of test of the work reported by Chi et al [13], we have considered structural families and fold types of protein (domains) as identified in this database. For exhaustive search we randomly selected maximum possible number of high-resolution structures of proteins the structural domains of which are authenticated in SCOP such that a comparable number of non-redundant observations are available from each of the four classes of interest and such that samples from each class will contain different possible sizes and orientation of the structural domain it represents.…”

“…Our earlier studies on ab initio prediction of protein tertiary structure using only the primary sequence have shown this feature as a statistically significant variable in correlation of the interresidue distances in primary and tertiary structures of proteins [10,14]. Moreover, inclusion of this feature does not increase the complexity of computing the feature-vector, so we have included it along with the features used by Chi et al [12,13].…”

“…Chi et al [12,13] have used 25-dimensional feature vector for fast protein structure retrieval and fold classification. We have attempted structural classification at the first level of the hierarchy in SCOP considering the local and global quantitative features used by them.…”

Ligands in PSI structures

“…Consequently, many of the proteins solved have little or no previous experimental characterization and have been classified as domains of unknown function (DUFs; Bateman et al, 2010) or have only a tentative functional annotation based on amino-acid sequence homology. A variety of online tools and web-based search engines, such as EBI-SSM (Krissinel & Henrick, 2004), DALI (Holm et al, 2008), VAST (Gibrat et al, 1996) and fast-SCOP (Chi et al, 2006), allow the inference of function based on structural similarity. However, these approaches have their limitations.…”