Classification of conformational stability of protein mutants from 2D graph representation of protein sequences using support vector machines

Easy Identification of Residues Involved on Structural Differences Between Nonphosphorylated and Phosphorylated CDK2Cyclin A Complexes Using Two-Dimensional Networks

The structures of proteins in Protein Data Bank (PDB) contain a lot of information that can be revealed through the use of tools to facilitate their organization and analysis. The increase in available structural data of nonphosphorylated and phosphorylated CDK2cyclin A (npCDK2cycA and pCDK2cycA) complexes has enabled a more realistic description of the fine structural details of the interface residues of these proteins. This work reports the application of two-dimensional network representations (TDNRs) to the structures deposited in PDB to distinguish the differences in the surface between both complexes due to phosphorylation. As a result, a detailed map of the hydrogen bonds (HBs) and hydrophobic interactions between the T-loop residues of CDK2 and the residues of cycA that are different among nonphosphorylated and phosphorylated complexes were described. In addition, we found some interesting subtle differences in the CDK2cycA interface between nonphosphorylated and phosphorylated complexes due to residues that are not located at the T-loop of CDK2. We noted that some HB interactions in CDK2cycA complex are reinforced when the CDK2 is phosphorylated.

Classification of Plasmodium falciparum glucose-6-phosphate dehydrogenase inhibitors by support vector machine

Plasmodium falciparum glucose-6-phosphate dehydrogenase (PfG6PD) has been considered as a potential target for severe forms of anti-malaria therapy. In this study, several classification models were built to distinguish active and weakly active PfG6PD inhibitors by support vector machine method. Each molecule was initially represented by 1,044 molecular descriptors calculated by ADRIANA.Code. Correlation analysis and attribute selection methods in Weka were used to get the best reduced set of molecular descriptors, respectively. The best model (Model 2w) gave a prediction accuracy (Q) of 93.88 % and a Matthew's correlation coefficient (MCC) of 0.88 on the test set. Some properties such as [Formula: see text] atom charge, [Formula: see text] atom charge, and lone pair electronegativity-related descriptors are important for the interaction between the PfG6PD and the inhibitor.

Classification of voltage-gated K(+) ion channels from 3D pseudo-folding graph representation of protein sequences using genetic algorithm-optimized support vector machines

Voltage-gated K(+) ion channels (VKCs) are membrane proteins that regulate the passage of potassium ions through membranes. This work reports a classification scheme of VKCs according to the signs of three electrophysiological variables: activation threshold voltage (V(t)), half-activation voltage (V(a50)) and half-inactivation voltage (V(h50)). A novel 3D pseudo-folding graph representation of protein sequences encoded the VKC sequences. Amino acid pseudo-folding 3D distances count (AAp3DC) descriptors, calculated from the Euclidean distances matrices (EDMs) were tested for building the classifiers. Genetic algorithm (GA)-optimized support vector machines (SVMs) with a radial basis function (RBF) kernel well discriminated between VKCs having negative and positive/zero V(t), V(a50) and V(h50) values with overall accuracies about 80, 90 and 86%, respectively, in crossvalidation test. We found contributions of the "pseudo-core" and "pseudo-surface" of the 3D pseudo-folded proteins to the discrimination between VKCs according to the three electrophysiological variables.