Current Computer - Aided Drug Design - Volume 11, Issue 4, 2015

Two and Three-dimensional quantitative structure-activity relationship (2D, 3D-QSAR) study was performed for some pyrazole-thiazolinone derivatives as EGFR kinase inhibitors using the CoMFA, CoMSIA and GA-MLR methods. The utilized data set was split into training and test set based on hierarchical clustering technique. From the five CoMSIA descriptors, electrostatic field presented the highest correlation with the activity. The statistical parameters for the CoMFA (r2=0.862, q2=0.644) and CoMSIA (r2=0.851, q2=0.740) were obtained for the training set with the common substructure-based alignment. The obtained parameters indicated the superiority of the CoMSIA model over the CoMFA model. A test set consisted of seven compounds was used to evaluate the proposed models. The results of contour maps which were presented by each method lead to some insights for increasing the inhibition activity of compounds. The 2D-QSAR model was built based on three descriptors selected by genetic algorithm and showed high predictive ability (R2 train= 0.843, Q2 LOO=0.787). Molecular docking study was also performed to understand the type interactions presented in binding site of the receptor and ligand. The developed models in parallel with molecular docking can be employed to design and derive novel compounds with the potent EGFR inhibitory activity.

The PI3K pathway is a signal transduction process including oncogenes and receptor tyrosine kinase regulating cellular functions i.e., survival, protein synthesis, and metabolism. In the present work, we have investigated the role of water molecules on inhibitor’s binding orientation in crystal structures of PI3K pathway targets using molecular docking approach. AutoDock v4.2 docking software was employed to dock PI3Kγ and its known inhibitors viz., wortmannin, quercetin, myricetin and pyridyl-triazine. Besides, serpentine was also docked on the same binding pocket, subsequently its anticancer activity was evaluated through in vitro experiment. Docking studies have been performed in the presence as well as in absence of water molecules at the binding pocket, and results were compared with crystallographic structural data. The comparison was done on the basis of binding energy, RMSD, inhibition constant (Ki), conserved and bridging water molecules, and found that, while considering water molecules during docking experiments, it increases the binding affinity of PI3K inhibitors.

A series of cellular automata models of bacteria were created, where encounters with models of the immune system and a model of an antibiotic drug were present. The dose of the antibiotic, its potency and the timeliness of its administration were variables. The emergence of antibiotic resistance by the bacteria was an outcome associated with the administration of the drug. The models created in these studies were found to closely relate to clinical experiences, making the general model useful for further simulation studies.

A short-lived membrane protein IRHOM2 pedals a cascade of events by regulating Epidermal Growth Factor Receptor (EGFR) signalling in parallel with metalloproteases which results their involvement in cancer as well as in rheumatoid arthritis. Therefore, IRHOM2 is a potential therapeutic drug target for these diseases, but its 3D-structure has not been reported yet. In this study, the three-dimensional structure of the IRHOM2 protein was generated using I-TASSER (Iterative Threading Assembly Refinement) server. The modeled structure of IRHOM2 receptor was validated using various Structural Analysis and Verification Server (SAVES) in which 99.7% of amino acid residues are present in the favoured regions of the Ramachandran Plot. Further, the refined modeled structure was subjected to molecular dynamics simulation & docking analysis. Virtual screening studies were carried out using Glide with various selective libraries containing 24552 compounds and the analysis indicated extensive hydrogen bonding network and hydrophobic interactions which play a significant role in its binding. Docking results were analyzed for high ranking compounds using a consensus based docking score to calculate the binding affinity as a measure of protein–ligand interactions. The top ranking molecule against IRHOM2 active site has a glide g-score of -12.565 kcal/mol and glide e-model score of -74.967 with 3 hydrogen bonds and 11 hydrophobic contacts. This compound may act as probable inhibitor against these chronic diseases but further in vitro studies are required.

In this paper, molecular topology was used to develop a mathematical model capable of classifying compounds according to their antibacterial activity. Topological indices were used as structural descriptors and their relation to antibacterial activity was determined by applying linear discriminant analysis (LDA) on a group of quinolones, widely used nowadays because of their broad spectrum of activity, well tolerance profile and advantageous pharmacokinetic properties. The topological model of activity obtained included two discriminant functions, selected by a combination of various statistical paremeters such as Fisher-Snedecor F and Wilk’s lambda, and allows the reliable prediction of antibacterial activity in any organic compound. After a virtual pharmacological screening on a library of 6375 compounds, the model has selected 263 as active compounds, from which 40% have proven antibacterial activity. The results obtained clearly reveal the high efficiency of molecular topology for the prediction of pharmacological activities. These models are very helpful in the discovery of new applications of natural and synthetic molecules with different chemical or biological properties. Therefore, we finally present 158 strong candidates to be developed as novel antibacterials.

Flap endonuclease-I (FEN-1) is involved in DNA repair and considered to be a novel target for the development of anticancer agents. N-hydroxy urea derivatives have been reported as FEN-1 inhibitors. To derive in vitro and in silico correlation, we have performed 2D-quantitative structure activity relationship (QSAR) analysis and docking studies on these compounds. 2D-QSAR models were developed using multiple linear regression (MLR) analysis and cross-validation using leave one out (LOO) method. The best model displayed R2 of 0.806 and Q2 of 0.607. Docking study revealed key interactions with desired amino acids and compare well with the in vitro potency of the reported compounds. Both studies reveal a link between FEN-1 inhibition and physicochemical descriptors or interactions with amino acids in active site. The information generated is first of its kind and may be helpful in the design of novel FEN-1 inhibitors.

A study of 30 flavonoid derivatives, taken from PubChem database and docked on flavonoid 3-O-glucosyltransferase 3HBF, next submitted to a QSAR study, performed within a hypermolecule frame, to model their LD50 values, is reported. The initial set of molecules was split into a training set and the test set (taken from the best scored molecules in the docking test); the predicted LD50 values, computed on similarity clusters, built up for each of the molecules of the test set, surpassed in accuracy the best model. The binding energies to 3HBF protein, provided by the docking step, are not related to the LD50 of these flavonoids, more protein targets are to be investigated in this respect. However, the docking step was useful in choosing the test set of molecules.

Despite being identified as the most potent receptor related to vasoconstriction, human urotensin-II receptor (hUT) has not been fully explored as a target for the treatment of cardiovascular diseases. In view of this and with an aim to identify precise structural requirements for binding of hUT antagonists, we endeavoured to develop, for the first time, multivariate QSAR models using chemometric methods like partial least squares (PLS) and feed-forward neural network (FFNN). A set of 48 pyrrolidine derivatives having hUT binding affinity was used for multivariate model development. The accuracy and predictability of the developed models was evaluated using crossvalidation. The PLS model showed good correlation between selected descriptors and Ki values (r2 =0.745 and r2 (CV) =0.773). However, the predictive performance of FFNN was better than the PLS technique with r2 =0.810. The study clearly suggests the role of lipophilic and steric descriptors in the ligand-hUT interactions. The QSAR models generated can be successfully extended to predict the binding affinities and for the effective design of novel hUT antagonists.