Field- and Gaussian-based 3D-QSAR studies on barbiturate analogs as MMP-9 inhibitors

Matrix metalloproteinase-9 (MMP-9) is one of the important enzymes belongs to gelatinase family and involved in multiple cellular processes including proliferation, angiogenesis, and metastasis. Various studies show that N-substituted homopiperazine barbiturates are promising and also selective for the MMP-9 inhibition. In this study, we selected and reported 49 barbiturate derivatives as inhibitors of MMP-9 and performed structure-based 3D-QSAR studies to elucidate the important structural features responsible for binding affinity. Crystal structure of MMP-9 complexed with barbiturate available in PDB was selected and performed quantum-polarized ligand docking for the reported inhibitors. Receptor-based alignment obtained from the docked poses were used for field- and Gaussian-based 3D-QSAR study. The results of both field and Gaussian models gave the good predictive correlation coefficient (test set) q ² of 0. 77 and 0.85, and the conventional correlation coefficient (training set) r ² of 0.845 and 0.928, respectively. The results of 3D-QSAR and docking studies validate each other and provided insight into the structural requirements for MMP-9 inhibition. The outcomes of the contour maps for both steric and Gaussian models results indicate that steric and hydrophobic interactions are major contributing factor for the activity and these findings will be useful to design the new inhibitors against MMP-9.     

Prospective atom-based 3D-QSAR model prediction,pharmacophore generation,and molecular docking study of carbamate derivatives as dual inhibitors of AChE and MAO-B for Alzheimer’s disease

Carbamates are well known for AChE as well as MAO inhibition. In this study, atom-based 3D-QSAR model generation, virtual screening, and molecular docking studies were performed for a known series of 31 carbamate derivatives. The best hypothesis yielded four different pharmacophoric features with statistically significant 3D-QSAR model (correlation coefficient of R ² = 0.994 for training set molecules and very good predictive powers with Q ² and Pearson-R were 0.60 and 0.91, respectively). By virtual screening done against Schrödinger database, we identified 11 distinct drug-like molecules binding to both targets AChE and MAO-B efficiently. This generated 3D-QSAR hybrid model for dual enzymes provides basis for new structural scaffold would serve as building blocks in designing drug-like molecules for Alzheimer’s disease.     

Structural features of GABAA receptor antagonists: pharmacophore modeling and 3D-QSAR studies

Pharmacophore modeling, comparative molecular field analysis (CoMFA), and comparative molecular similarity indices analysis (CoMSIA) studies have been carried out on 5-(4-piperidyl)-3-isoxazolol (4-PIOL) analogs as GABA_A receptor antagonists in this study. The best pharmacophore hypothesis generated by PHASE was ADHPR.6, which comprised a hydrogen bond acceptor (A), a hydrogen bond donor (D), a hydrophobic group (H), a positively charged group (P), and an aromatic ring (R). The pharmacophore model provided a good alignment for the further 3D-QSAR analyses, which presented a good R ² value of 0.943, 0.930, and 0.916 for atom-based QSAR model, CoMFA model, and CoMSIA model, respectively. All QSAR models presented good statistical significance and predictivity, the corresponding Q ² values for each 3D-QSAR model are 0.794, 0.569, and 0.637, respectively. Both pharmacophore and CoMSIA results showed that the hydrophobic sites are the key structural feature for GABA_A receptor antagonists with high activities.     

Design of novel focal adhesion kinase inhibitors using 3D-QSAR and molecular docking

Focal adhesion kinase (FAK) plays a primary role in regulating the activity of many signaling molecules. Increased FAK expression has been implicated in a series of cellular processes, including cell migration and survival. Inhibiting the activity of FAK for cancer therapy is currently under investigation. Hence, FAK and its inhibitors have been the subject of intensive research. To understand the structural factors affecting inhibitory potency, molecular docking and 3D-QSAR modeling were studied in this project. CoMFA and CoMSIA methods were used for deriving 3D-QSAR models, which were trained with 78 compounds and then were evaluated for predictive ability with additional 19 compounds. Two substructure-based 3D-QSAR models, including CoMFA model (r ² = 0.930; q ² = 0.629) and CoMSIA model (r ² = 0.952; q ² = 0.586), had a good quality to predict the biological activities of new compounds. Meanwhile, using the flexible docking strategy, two docking-based 3D-QSAR models (CoMFA with r ² = 0.914; q ² = 0.594; CoMSIA with r ² = 0.914; q ² = 0.524) were also constructed. The structure–activity relationship has been illustrated clearly by the contour maps gained from the 3D-QSAR models in combination with the docked binding structures. All the results indicated that it might be useful in the rational design of novel FAK inhibitors.     

Pharmacophore modeling, 3D-QSAR, and molecular docking study on naphthyridine derivatives as inhibitors of 3-phosphoinositide-dependent protein kinase-1

The 3-phosphoinositide-dependent protein kinase-1 (PDK1) is an imminent target for discovering novel anticancer drugs. In order to understand the structure–activity correlation of naphthyridine-based PDK-1 inhibitors, we have carried out a combined pharmacophore, three-dimensional quantitative structure–activity relationship (3D-QSAR), and molecular docking studies. The study has resulted in six point pharmacophore models with four hydrogen bond acceptors (A), one hydrogen bond donor (D), and one aromatic ring (R) are used to derive a predictive atom-based 3D-QSAR model. The generated 3D-QSAR model shows that the alignment has good correlation coefficient for the training set compounds which comprises the values of R ² = 0.96, SD = 0.2, and F = 198.2. Test set compounds shows Q ² = 0.84, RMSE = 0.56, and Pearson-R = 0.84. The external validation was carried out to validate the predicted QSAR model which shows good predictive power of $ r_{m}^{2} $  = 0.83 and k = 1.01, respectively. The external validation results also confirm the fitness of the model. The results indicated that, atom-based 3D-QSAR models and further modifications in PDK1 inhibitors via pharmacophore hypothesis are rational for the prediction of the activity of new inhibitors in prospect of drug design.     

3D quantitative structure–activity relationship for quinoline, benzimidazole and benzofuran-based analogs as phosphodiesterases IV (PDE-IV) inhibitors

PDE-IV is one of the important targets in the treatment of asthma, COPD, and rheumatoid arthritis. In the search for novel PDE-IV inhibitors a 3D-QSAR study was performed on PDE-IV inhibitors by means of pharmacophore mapping using PHASE, Schrödinger-9. The 3D-QSAR obtained from AAHHRR-1024 hypothesis was found to be statistically significant with r ² = 0.9766 and q ² = 0.8759 with 7 PLS factors. The statistical significance of the model was confirmed by a very low value of RMSE 0.4795. The "Pearson-R" value of 0.9376 suggests a very good predictive ability of the hypothesis generated. The present study demonstrates a robust 3D-QSAR model of PDE-IV inhibitor with the help of AAHHRR-1024 hypothesis, which will help in designing novel inhibitors.     

Pharmacophore-based 3D-QSAR study of fungal chitin synthase inhibitors

To design new compounds with enhanced activity against the fungal chitin synthase enzyme, 3D-pharmacophore models were generated and QSAR study was carried out on 44 novel homoallylamines and related compounds, nikkomycin, maleimide, chalcones, and quinolin-2-one derivatives. A three-point pharmacophore with two hydrophobic (H) and one aromatic ring (R) as pharmacophore features was developed by PHASE module of Schrodinger molecular modeling suite. The pharmacophore hypothesis yielded a statistically significant 3D-QSAR model, with a correlation coefficient of R ² of 0.84 for training set compounds. The model generated showed excellent predictive power, with a correlation coefficient of Q ² of 0.63 and Pearson-R value of 0.82 for a randomly chosen test set of nine compounds. The 3D-QSAR model explains the structure–activity relationship of these compounds which may help in the design and development of novel fungal chitin synthase inhibitors.     

3D-QSAR of amino-substituted pyrido[3,2B]pyrazinones as PDE-5 inhibitors

A 3D-QSAR study on amino-substituted pyrido[3,2b]pyrazinones as PDE-5 inhibitors was successfully performed by means of pharmacophore mapping using PHASE module of Schrödinger-9. The 3D-QSAR obtained from AADHRR-183 hypothesis was found to be statistically good with r ² = 0.95 and q ² = 0.81 taking PLS factor 4. The statistical significance of the model was also confirmed by a high value of Fisher ratio of 85.1 and a very low value of RMSE 0.29. One of the other parameters which signify the model predictivity is Pearson R. Its value of 0.91 shows that the correlation between predicted and observed activities for the test set compounds is excellent. Hydrophobic groups are important for PDE-5 inhibition while H-bond donor groups are less favorable for the same. Electron withdrawing groups are favorable if include at ring A in the structures while unfavorable at other sites. Thus, it can be assumed that the present QSAR analysis is enough to demonstrate PDE-5 inhibition with the help of AADHRR-183 hypothesis and will help in designing novel and potent PDE-5 inhibitors.     

QSAR and docking studies of novel β-carboline derivatives as anticancer

Quantitative structure–activity relationship (QSAR) studies were performed on β-carboline derivatives for prediction of anticancer activity. The statistically significant 2D-QSAR model having r ² = 0.726 and q ² = 0.654 with pred_r ² = 0.763 was developed by stepwise multiple linear regression method. In order to understand the structural requirement of these β-carboline derivatives, a ligand-based pharmacophore 3D-QSAR model was developed. The five-point pharmacophore hypothesis yielded a 3D-QSAR model with good partial least-square statistics results (r ² = 0.73, Q _ext ²  = 0.755, F = 67.5, SD = 0.245, RMSE = 0.241, Pearson-R = 0.883). A docking study revealed the binding orientations of these derivatives at the active site amino residues of DNA intercalate (PDB ID: 1D12). The results of 2D-QSAR, atom-based 3D-QSAR, and docking studies gave detailed structural insights as well as highlighted important binding features of β-carboline derivatives as anticancer agent which provided guidance for the rational design of novel potent anticancer agents.     

Molecular modeling study of HIV-1 gp120 attachment inhibitors

The viral glycoprotein 120 (gp120) is a glycoprotein exposed on viral surface. The gp120 is essential for virus entry into cells as it plays a vital role in seeking out specific cell surface receptors for entry. In this article, we performed docking and three-dimensional quantitative structure activity relationship (3D-QSAR) study on a series of 48 indole glyoxamide derivatives as gp120 inhibitors. Docking study revealed that the inhibitor docked deeply into the gp120 cavity rather than Phe43 of cluster of differentiation 4 (CD4). 3D-QSAR methodologies, comparative molecular field analysis (CoMFA), and comparative molecular similarity indices analysis (CoMSIA) were utilized to rationalize the structural variations with their inhibitory activities. The docked pose of the most potent molecule (43) was used to determine the structures of other molecules. The CoMFA yielded a model with cross-validated correlation coefficient of (q ²) 0.73 and non-cross-validated correlation coefficient of (r ²) 0.89 with optimum number of components (N?=?3). The CoMSIA models were obtained with the combination of various parameters. Final model was computed with steric, hydrophobic- and hydrogen-bond acceptor (SHA) parameters with reasonable statistics (q ²?=?0.80, r ²?=?0.94 and N?=?5). The predictive power of developed CoMFA and CoMSIA models were assessed by test set (nine molecules). The predictive r _pred ² for CoMFA and CoMSIA model was found to be 0.93 and 0.74, respectively. The generated contour maps were plotted onto the gp120 active site to correlate structural variations with their biological activity in protein environment. Contour map analyses showed the importance of 4-F substitution of indole ring, which made essential electronic interaction with the crucial residue (Trp427). The 3D models could explain nicely the structure–activity relationships of indole glyoxamide analogs. This would give proper guidelines to further enhance the activity of novel inhibitors.     

Three-dimensional quantitative structure activity relationship analysis of anilinoquinazolines for c-Src kinase inhibition

Three-dimensional quantitative structure activity relationship (3D-QSAR) models was developed using molecular field analysis (MFA) for 36 anilinoquinazoline derivatives, inhibiting c-Src kinase. The QSAR model was developed using 29 compounds and its predictive ability was assessed using a test set of seven compounds. The predictive 3D-QSAR model has conventional r ² values of 0.961 while the cross-validated coefficient q ² and bootstrap correlation coefficient r _BS² values of 0.910 and 0.957, respectively. The developed model provides a powerful tool to design potent c-Src inhibitors as novel antitumor agents. Six new inhibitors were designed and their pIC₅₀ were predicted.     

Computation of pharmacophore models for the prediction of mitogen-activated protein kinase activated protein kinase-2 inhibitory activity of pyrrolopyridines

This article is an attempt to formulate the three-dimensional pharmacophore modelling of pyrrolopyridine derivatives inhibiting mitogen-activated protein kinase activated protein kinase-2 (MK₂). To understand the essential structural features for MK₂ inhibitors, pharmacophore hypothesis were built on the basis of a set of known MK₂ inhibitors selected from literature using PHASE program. Three pharmacophore models with one hydrogen-bond acceptor (A), two hydrogen-bond donors (D), one hydrophobic group (H) and one aromatic ring (R) as pharmacophoric features were developed. Amongst them the pharmacophore hypothesis ADDHR1 yielded a statistically significant 3D-QSAR model with 0.926 as R ² value and was considered to be the best pharmacophore hypothesis. The developed pharmacophore model was externally validated by predicting the activity of test set molecules. The squared predictive correlation coefficient of 0.882 was observed between experimental and predicted activity values of test set molecules. The geometry and features of pharmacophore was expected to be useful for the design of selective MK₂ inhibitors.     

QSAR studies for some thiazolidine-2,4-dione derivatives as PIM-2 kinase inhibitors

Quantitative structure–activity relationship (QSAR) studies were performed on a series of 21 thiazolidine-2,4-dione derivatives to find the structural requirements for PIM-2 kinase inhibitory activity by two-dimensional (2D-QSAR), group-based (G-QSAR) and three-dimensional (3D-QSAR) studies. In the present study, widely used technique viz. stepwise forward–backward (SW-FB) has been applied for the development of 2D- and G-QSAR as variable selection method. The statistically significant best 2D-QSAR model was developed by partial least squares regression (PLSR) having r ² = 0.78, q ² = 0.63 with pred_r ² = 0.78. The statistically significant best G-QSAR model was developed by PLSR method having r ² = 0.89, q ² = 0.79 and pred_r ² = 0.82. The 3D-QSAR studies were performed by k-nearest neighbor molecular field analysis along with genetic algorithm method which showed q ² = 0.64 and pred_r ² = 0.94. A docking study revealed the binding orientations of these inhibitors at active site amino acid residues (PHE 43, ASP 124, ASP 182 and GLU 83) of PIM-2 enzyme (PDB ID: 3IWI). The results of this study may be useful to (medicinal) chemists to design more potent thiazolidine-2,4-dione analogs as PIM-2 kinase inhibitors.     

A combined 3D QSAR and pharmacophore-based virtual screening for the identification of potent p38 MAP kinase inhibitors: an in silico approach

p38 kinase plays a vital role in inflammation mediated by tumor necrosis factor-α and interleukin-1β pathways. Inhibition of p38 kinase provides an effective way to treat inflammatory diseases. 3D-QSAR study was performed to obtain reliable comparative molecular field analysis (CoMFA) and comparative molecular similarity indices analysis (CoMSIA) models for a series of p38 inhibitors with three different alignment methods (Receptor based, atom by atom matching, and pharmacophore based). Among the different alignment methods, better statistics were obtained with receptor-based alignment (CoMFA: q ² = 0.777, r ² = 0.958; CoMSIA: q ² = 0.782, r ² = 0.927). Superposing CoMFA/CoMSIA contour maps on the p38 active site gave a valuable insight to understand physical factors which are important for binding. In addition, this pharmacophore model was used as a 3D query for virtual screening against NCI database. The hit compounds were further filtered by docking and scoring, and their biological activities were predicted by CoMFA and CoMSIA models.     

Molecular modeling studies on series of Btk inhibitors using docking,structure-based 3D-QSAR and molecular dynamics simulation: a combined approach

Bruton tyrosine kinase (Btk) is a non-receptor tyrosine kinase. It is a crucial component in BCR pathway and expressed only in hematopoietic cells except T cells and Natural killer cells. BTK is a promising target because of its involvement in signaling pathways and B cell diseases such as autoimmune disorders and lymphoma. In this work, a combined molecular modeling study of molecular docking, 3D-QSAR and molecular dynamic (MD) simulation were performed on a series of 2,5-diaminopyrimidine compounds as inhibitors targeting Btk kinase to understand the interaction and key residues involved in the inhibition. A structure based CoMFA (q ² = 0.675, NOC = 5, r ² = 0.961) and COMSIA (q ² = 0.704, NOC = 6, r ² = 0.962) models were developed from the conformation obtained by docking. The developed models were subjected to various validation techniques such as leave-five-out, external test set, bootstrapping, progressive sampling and rm ² metrics and found to have a good predictive ability in both internal and external validation. Our docking results showed the important residues that interacts in the active site residues in inhibition of Btk kinase. Furthermore, molecular dynamics simulation was employed to study the stability of the docked conformation and to investigate the binding interactions in detail. The MD simulation analyses identified several important hydrogen bonds with Btk, including the gatekeeper residue Thr474 and Met477 at the hinge region. Hydrogen bond with active site residues Leu408 and Arg525 were also recognized. A good correlation between the MD results, docking studies and the contour map analysis are observed. This indicates that the developed models are reliable. Our results from this study can provide insights in the designing and development of more potent Btk kinase inhibitors.     

Quasi 4D-QSAR and 3D-QSAR study of the pan class I phosphoinositide-3-kinase (PI3K) inhibitors

The class I phosphoinositide-3-kinases (PI3Ks) is currently investigated and attracted as a promising target toward anticancer therapies. The quasi 4D-QSAR model is developed by a training set of 30 pan class I PI3K inhibitors. This methodology is based on the generation of a conformational ensemble profile for each compound instead of only one conformation, followed by the calculation of intermolecular interaction energies at each grid point considering probes and all aligned conformations resulting from molecular dynamic simulations. A comparison of the proposed methodology with comparative molecular field analysis (CoMFA) formalism has been performed. This paradigm explores jointly the main features of CoMFA and 4D-QSAR models. The best 4D-QSAR model is checked for free from chance correlation, reliability and robustness by leave-N-out cross-validation and Y-randomization in addition to analysis of the independent test set. Statistical parameters of the best 4D-QSAR model are R ² = 0.871, q _LOO ²  = 0.661, and R _Pred ²  = 0.751. The results of the suggested model are in good agreement with docking study that was previously reported by Rewcastle et al. (J Med Chem 54:7105–7126, 2011).