Support vector machines in HTS data mining: Type I MetAPs inhibition study

This article reports a successful application of support vector machines (SVMs) in mining high-throughput screening (HTS) data of a type I methionine aminopeptidases (MetAPs) inhibition study. A library with 43,736 small organic molecules was used in the study, and 1355 compounds in the library with 40% or higher inhibition activity were considered as active. The data set was randomly split into a training set and a test set (3:1 ratio). The authors were able to rank compounds in the test set using their decision values predicted by SVM models that were built on the training set. They defined a novel score PT50, the percentage of the test set needed to be screened to recover 50% of the actives, to measure the performance of the models. With carefully selected parameters, SVM models increased the hit rates significantly, and 50% of the active compounds could be recovered by screening just 7% of the test set. The authors found that the size of the training set played a significant role in the performance of the models. A training set with 10,000 member compounds is likely the minimum size required to build a model with reasonable predictive power.     

A combined approach of docking and 3D QSAR study of beta-ketoacyl-acyl carrier protein synthase III (FabH) inhibitors

The enzyme FabH catalyzes the initial step of fatty acid biosynthesis via a type II fatty acid synthase. The pivotal role of this essential enzyme combined with its unique structural features and ubiquitous occurrence in bacteria has made it an attractive new target for the development of antibacterial and antiparasitic compounds. Three-dimensional quantitative structure-activity relationship (3D QSAR) studies such as comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) and docking simulations were conducted on a series of potent benzoylaminobenzoic acids. Docking studies were employed to position the inhibitors into the FabH active site to determine the probable binding conformation. A reasonable correlation between the predicated binding free energy and the inhibitory activity was found. CoMFA and CoMSIA were performed based on the docking conformations, giving q(2) of 0.637 and 0.697 for CoMFA and CoMSIA models, respectively. The predictive ability of the models was validated using a set of compounds that were not included in the training set and progressive scrambling test. Mapping the 3D QSAR models to the active site of FabH related that some important amino acid residues are responsible for protein-inhibitor interaction. These results should be applicable to the prediction of the activities of new FabH inhibitors, as well as providing structural understanding.     

Molecular docking and 3D-QSAR studies of Yersinia protein tyrosine phosphatase YopH inhibitors

Three-dimensional quantitative structure-activity relationship (QSAR) studies were conducted on two classes of recently explored compounds with known YopH inhibitory activities. Docking studies were employed to position the inhibitors into the YopH active site to determine the probable binding conformation. Good correlations between the predicated binding free energies and the inhibitory activities were found for two subsets of phosphate mimetics: alpha-ketocarboxylic acid and squaric acid (R2=0.70 and 0.68, respectively). The docking results also provided a reliable conformational alignment scheme for 3D-QSAR modeling. Comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) were performed based on the docking conformations, giving q2 of 0.734 and 0.754 for CoMFA and CoMSIA models, respectively. The 3D-QSAR models were significantly improved after removal of an outlier (q2=0.829 for CoMFA and q2=0.837 for CoMSIA). The predictive ability of the models was validated using a set of compounds that were not included in the training set. Mapping the 3D-QSAR models to the active site of YopH provides new insight into the protein-inhibitor interactions for this enzyme. These results should be applicable to the prediction of the activities of new YopH inhibitors, as well as providing structural implications for designing potent and selective YopH inhibitors as antiplague agents.     

3D QSAR studies on T-type calcium channel blockers using CoMFA and CoMSIA

Comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) were performed on a series of isoxazolyl compounds as a potent T-type calcium channel blockers. A set of 24 structurally similar compounds served to establish the model. Four different conformations of the most active compound were used as template structures for the alignment, three of which were obtained from Catalyst pharmacophore modeling and one by using SYBYL random search option. All CoMFA and CoMSIA models gave cross-validated r(2) (q(2)) value of more than 0.5 and conventional r(2) value of more than 0.85. The predictive ability of the models was validated by an external test set of 10 compounds, which gave satisfactory pred r(2) values ranging from 0.577 to 0.866 for all models. Best predictions were obtained with CoMFA std model of Conformer no: 3 alignment (q(2)=0.756, r(2)=0.963), giving predictive r(2) value of 0.866 for the test set. CoMFA and CoMSIA contour maps were used to analyze the structural features of the ligands accounting for the activity in terms of positively contributing physicochemical properties: steric, electrostatic, hydrophobic and hydrogen bonding fields.     

Molecular docking and 3D-QSAR studies on inhibitors of DNA damage signaling enzyme human PARP-1

Poly (ADP-ribose) polymerase-1 (PARP-1) operates in a DNA damage signaling network. Molecular docking and three dimensional-quantitative structure activity relationship (3D-QSAR) studies were performed on human PARP-1 inhibitors. Docked conformation obtained for each molecule was used as such for 3D-QSAR analysis. Molecules were divided into a training set and a test set randomly in four different ways, partial least square analysis was performed to obtain QSAR models using the comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA). Derived models showed good statistical reliability that is evident from their r2, q2(loo) and r2(pred) values. To obtain a consensus for predictive ability from all the models, average regression coefficient r2(avg) was calculated. CoMFA and CoMSIA models showed a value of 0.930 and 0.936, respectively. Information obtained from the best 3D-QSAR model was applied for optimization of lead molecule and design of novel potential inhibitors.     

Molecular docking and 3D-QSAR studies on β-phenylalanine derivatives as dipeptidyl peptidase IV inhibitors

Three-dimensional quantitative structure-activity relationship (3D-QSAR) and molecular docking studies were carried out to explore the binding of 73 inhibitors to dipeptidyl peptidase IV (DPP-IV), and to construct highly predictive 3D-QSAR models using comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA). The negative logarithm of IC₅₀ (pIC₅₀) was used as the biological activity in the 3D-QSAR study. The CoMFA model was developed by steric and electrostatic field methods, and leave-one-out cross-validated partial least squares analysis yielded a cross-validated value (r_cv² {\hbox{r}}_{{\rm{cv}}}^{\rm{2}} ) of 0.759. Three CoMSIA models developed by different combinations of steric, electrostatic, hydrophobic and hydrogen-bond fields yielded significant r_cv² {\hbox{r}}_{{\rm{cv}}}^{\rm{2}} values of 0.750, 0.708 and 0.694, respectively. The CoMFA and CoMSIA models were validated by a structurally diversified test set of 18 compounds. All of the test compounds were predicted accurately using these models. The mean and standard deviation of prediction errors were within 0.33 and 0.26 for all models. Analysis of CoMFA and CoMSIA contour maps helped identify the structural requirements of inhibitors, with implications for the design of the next generation of DPP-IV inhibitors for the treatment of type 2 diabetes.     

A model for identifying HERG K+ channel blockers

Acquired long QT syndrome (LQTS) occurs frequently as a side effect of blockade of cardiac HERG K(+) channels by commonly used medications. A large number of structurally diverse compounds have been shown to inhibit K(+) current through HERG. There is considerable interest in developing in silico tools to filter out potential HERG blockers early in the drug discovery process. We describe a binary classification model that combines a 2D topological similarity filter with a 3D pharmacophore ensemble procedure to discriminate between HERG actives and inactives with an overall accuracy of 82%, with false negative and false positive rates of 29% and 15%, respectively. This model should be generally applicable in virtual library counterscreening against HERG.     

In silico studies for the rational discovery of anticonvulsant compounds

Theoretical models to virtual screening and rational design of anticonvulsant compounds based on a topological sub-structural molecular design (TOSS-MODE) approach are developed. These models, developed on the basis of data sets of great structural variability, permit the classification of compounds as active/inactive anticonvulsants and predict the quantitative anticonvulsant potency of such compounds. The classification model is applied to a virtual screening of anticonvulsant compounds by analyzing a data set of molecules reported in the literature. More than 88% of them were well classified by the current model. Active and inactive fragments are identified by using the present approach. Some of the active fragments are identified in anticonvulsant molecules as potential pharmacophores and one of them is analyzed in detail. The three-dimensional (3-D) features of this fragment are investigated in a series of five anticonvulsant compounds. Some structure–anticonvulsant activity relationships are derived on the basis of the 3-D structure of this fragment and some findings reported in the literature that indicate that it is an important pharamacophore are outlined.     

The discovery of inhibitors of Fas-mediated cell death pathway using the combined computational method

In this study, pharmacophore and 3D-QSAR models were developed for analogues of 3-substituted-benzofuran-2-carboxylate as inhibitors of Fas-mediated cell death pathways. Our pharmacophore model has good correspondence with experimental results and can explain the variance in biological activities coherently with respect to the structure of the data set compounds. The predictive power for our synthesized compounds were 0.96 for the pharmacophore model, 0.58 for the comparative molecular field analysis (CoMFA) model, and 0.57 for the comparative molecular similarity analysis (CoMSIA) model.     

Key features for designing phosphodiesterase-5 inhibitors

Phosphodiesterase superfamily is the key regulator of 3',5'-cyclic guanosine monophosphate (cGMP) decomposition in human body. Phosphodiesterase-5 (PDE-5) inhibitors, sildenafil, vardenafil and tadalafil, are well known oral treatment for males with erectile dysfunction. To investigate the inhibitory effects of traditional Chinese medicine (TCM) compounds to PDE-5, we performed both ligand-based and structure-based studies on this topic. Comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) studies were conducted to construct three dimensional quantitative structure-activity relationship (3D-QSAR) models of series of known PDE-5 inhibitors. The predictive models had cross-validated, q(2), and non cross-validated coefficient, r(2), values of 0.791 and 0.948 for CoMFA and 0.724 and 0.908 for CoMSIA. These two 3D-QSAR models were used to predict activity of TCM compounds. Docking simulations were performed to further analyze the binding mode of training set and TCM compounds. A putative binding model was proposed based on CoMFA and CoMSIA contour maps and docking simulations; formation of pi-stacking, water bridge and specific hydrogen bonding were deemed important interactions between ligands and PDE-5. Of our TCM compounds, engeletin, satisfied our binding model, and hence, emerged as PDE-5 inhibitor candidate. Using this study as an example, we demonstrated that docking should be conducted for qualitative purposes, such as identifying protein characteristics, rather than for quantitative analyses that rank compound efficacy based on results of scoring functions. Prediction of compound activity should be reserved for QSAR analyses, and scoring functions and docking scores should be used for preliminary screening of TCM database (http://tcm.cmu.edu.tw/index.php).