Protein Classification by Matching 3D Structures

Efficient Approaches for Retrieving Protein Tertiary Structures

“…We have already published some of these approaches, while some of them are novel. In [30] and [31], we have already presented the protein voxelbased descriptor, the protein ray-based descriptor with uniform interpolation of the protein backbone and the Haar wavelet-based protein descriptor with decomposition to the last possible level. In this paper, we introduce the protein ray-based descriptor with nonuniform interpolation, five novel protein wavelet-based descriptors, and our approach named Matrix Alignment by Sequence Alignment within Sliding Window (MASASW).…”

“…In [38], several 3D descriptors are presented, such as silhouettebased, depth buffer-based, volume-based, and voxel-based feature vectors. In [30] and [31], we have made some modifications of them for extracting suitable protein descriptors. In the approach given in [32], the proteins are represented by their fractal dimension which represents the self-similarity of the protein structure.…”

“…We have adapted the method given in [38] to extract the protein voxel-based descriptor. Additionally, some features of the primary and secondary structure of the protein are considered, as in [42], thus forming better integrated descriptor [30], [31].…”

“…We use this algorithm for extracting the geometrical features of the protein. Additionally some features of the primary and secondary structure are considered, thus forming our integrated protein voxel-based descriptor [30], [31].…”

“…There are a huge number of algorithms for retrieving protein tertiary structures which extract some features from the alternative representatives of the polypeptide chain such as 3D Spline [28], [29], or features from the 3D coordinates of the amino acids [30], [31], [32], [33], [34], [35], [36], [37]. In this way, the protein structure is represented by a multidimensional feature vector.…”