QN-Docking: An innovative molecular docking methodology based on Q-Networks

In silico insights into the design of novel NR2B-selective NMDA receptor antagonists: QSAR modeling, ADME-toxicity predictions, molecular docking, and molecular dynamics investigations

Based on a structural family of thirty-two NR2B-selective N-Methyl-D-Aspartate receptor (NMDAR) antagonists, two phenylpiperazine derivatives labeled C37 and C39 were conceived thanks to molecular modeling techniques, as novel NMDAR inhibitors exhibiting the highest analgesic activities (of pIC50 order) against neuropathic pain, with excellent ADME-toxicity profiles, and good levels of molecular stability towards the targeted protein of NMDA receptor. Initially, the quantitative structure-activity relationships (QSARs) models were developed using multiple linear regression (MLR), partial least square regression (PLSR), multiple non-linear regression (MNLR), and artificial neural network (ANN) techniques, revealing that analgesic activity was strongly correlated with dipole moment, octanol/water partition coefficient, Oxygen mass percentage, electronegativity, and energy of the lowest unoccupied molecular orbital, whose the correlation coefficients of generated models were: 0.860, 0.758, 0.885 and 0.977, respectively. The predictive capacity of each model was evaluated by an external validation with correlation coefficients of 0.703, 0.851, 0.778, and 0.981 respectively, followed by a cross-validation technique with the leave-one-out procedure (CVLOO) with Q2cv of 0.785, more than Y-randomization test, and applicability domain (AD), in addition to Fisher's and Student's statistical tests. Thereafter, ten novel molecules were designed based on MLR QSAR model, then predicted with their ADME-Toxicity profiles and subsequently examined for their similarity to the drug candidates. Finally, two of the most active compounds (C37 and C39) were chosen for molecular docking and molecular dynamics (MD) investigations during 100 ns of MD simulation time in complex with the targeted protein of NMDA receptor (5EWJ.pdb).

QSAR, ADME-Tox, molecular docking and molecular dynamics simulations of novel selective glycine transporter type 1 inhibitors with memory enhancing properties

A structural class of forty glycine transporter type 1 (GlyT1) inhibitors, was examined using molecular modeling techniques. The quantitative structure-activity relationships (QSAR) technology confirmed that human GlyT1 activity is strongly and significantly affected by constitutional, geometrical, physicochemical and topological descriptors. ADME-Tox in-silico pharmacokinetics revealed that L28 and L30 ligands were predicted as non-toxic inhibitors with a good ADME profile and the highest probability to penetrate the central nervous system (CNS). Molecular docking results indicated that the predicted inhibitors block GlyT1, reacting specifically with Phe319, Phe325, Tyr123, Tyr 124, Arg52, Asp475, Ala117, Ala479, Ile116 and Ile483 amino acids of the dopamine transporter (DAT) membrane protein. These results were qualified and strengthened using molecular dynamics (MD) study, which affirmed that the established intermolecular interactions for (L28, L30-DAT protein) complexes remain perfectly stable along 50 ns of MD simulation time. Therefore, they could be strongly recommended as therapeutics in medicine to improve memory performance.

A comprehensive review on recent approaches for cancer drug discovery associated with artificial intelligence

Through the revolutionization of artificial intelligence (AI) technologies in clinical research, significant improvement is observed in diagnosis of cancer. Utilization of these AI technologies, such as machine and deep learning, is imperative for the discovery of novel anticancer drugs and improves existing/ongoing cancer therapeutics. However, building a model for complicated cancers and their types remains a challenge due to lack of effective therapeutics that hinder the establishment of effective computational tools. In this review, we exploit recent approaches and state-of-the-art in implementing AI methods for anticancer drug discovery, and discussed how advances in these applications need to be considered in the current cancer therapeutics. Considering the immense potential of AI, we explore molecular docking and their interactions to recognize metabolic activities that support drug design. Finally, we highlight corresponding strategies in applying machine and deep learning methods to various types of cancer with their pros and cons.

3D-QSAR, ADME-Tox In Silico Prediction and Molecular Docking Studies for Modeling the Analgesic Activity against Neuropathic Pain of Novel NR2B-Selective NMDA Receptor Antagonists

A new class of selective antagonists of the N-Methyl-D-Aspartate (NMDA) receptor subunit 2B have been developed using molecular modeling techniques. The three-dimensional quantitative structure–activity relationship (3D-QSAR) study, based on comparative molecular field analysis (CoMFA) and comparative molecular similarity index analysis (CoMSIA) models, indicate that steric, electrostatic and hydrogen bond acceptor fields have a key function in the analgesic activity against neuropathic pain. The predictive accuracy of the developed CoMFA model (Q2 = 0.540, R2 = 0.980, R2 pred = 0.613) and the best CoMSIA model (Q2 = 0.665, R2 = 0.916, R2 pred = 0.701) has been successfully examined through external and internal validation. Based on ADMET in silico properties, L1, L2 and L3 ligands are non-toxic inhibitors of 1A2, 2C19 and 2C9 cytochromes, predicted to passively cross the blood–brain barrier (BBB) and have the highest probability to penetrate the central nervous system (CNS). Molecular docking results indicate that the active ligands (L1, L2 and L3) interact specifically with Phe176, Glu235, Glu236, Gln110, Asp136 and Glu178 amino acids of the transport protein encoded as 3QEL. Therefore, they could be used as analgesic drugs for the treatment of neuropathic pain.

QSAR, ADMET In Silico Pharmacokinetics, Molecular Docking and Molecular Dynamics Studies of Novel Bicyclo (Aryl Methyl) Benzamides as Potent GlyT1 Inhibitors for the Treatment of Schizophrenia

Forty-four bicyclo ((aryl) methyl) benzamides, acting as glycine transporter type 1 (GlyT1) inhibitors, are developed using molecular modeling techniques. QSAR models generated by multiple linear and non-linear regressions affirm that the biological inhibitory activity against the schizophrenia disease is strongly and significantly correlated with physicochemical, geometrical and topological descriptors, in particular: Hydrogen bond donor, polarizability, surface tension, stretch and torsion energies and topological diameter. According to in silico ADMET properties, the most active ligands (L6, L9, L30, L31 and L37) are the molecules having the highest probability of penetrating the central nervous system (CNS), but the molecule 32 has the highest probability of being absorbed by the gastrointestinal tract. Molecular docking results indicate that Tyr124, Phe43, Phe325, Asp46, Phe319 and Val120 amino acids are the active sites of the dopamine transporter (DAT) membrane protein, in which the most active ligands can inhibit the glycine transporter type 1 (GlyT1). The results of molecular dynamics (MD) simulation revealed that all five inhibitors remained stable in the active sites of the DAT protein during 100 ns, demonstrating their promising role as candidate drugs for the treatment of schizophrenia.

A point cloud-based deep learning strategy for protein-ligand binding affinity prediction

There is great interest to develop artificial intelligence-based protein-ligand binding affinity models due to their immense applications in drug discovery. In this paper, PointNet and PointTransformer, two pointwise multi-layer perceptrons have been applied for protein-ligand binding affinity prediction for the first time. Three-dimensional point clouds could be rapidly generated from PDBbind-2016 with 3772 and 11 327 individual point clouds derived from the refined or/and general sets, respectively. These point clouds (the refined or the extended set) were used to train PointNet or PointTransformer, resulting in protein-ligand binding affinity prediction models with Pearson correlation coefficients R = 0.795 or 0.833 from the extended data set, respectively, based on the CASF-2016 benchmark test. The analysis of parameters suggests that the two deep learning models were capable to learn many interactions between proteins and their ligands, and some key atoms for the interactions could be visualized. The protein-ligand interaction features learned by PointTransformer could be further adapted for the XGBoost-based machine learning algorithm, resulting in prediction models with an average Rp of 0.827, which is on par with state-of-the-art machine learning models. These results suggest that the point clouds derived from PDBbind data sets are useful to evaluate the performance of 3D point clouds-centered deep learning algorithms, which could learn atomic features of protein-ligand interactions from natural evolution or medicinal chemistry and thus have wide applications in chemistry and biology.