Current Computer - Aided Drug Design - Volume 15, Issue 2, 2019

Introduction: The development of a novel drug is an extremely complicated process that includes the target identification, design and manufacture, and proper therapy of the novel drug, as well as drug dose selection, drug efficacy evaluation, and adverse drug reaction control. Due to the limited resources, high costs, long duration, and low hit-to-lead ratio in the development of pharmacogenetics and computer technology, machine learning techniques have assisted novel drug development and have gradually received more attention by researchers. Methods: According to current research, machine learning techniques are widely applied in the process of the discovery of new drugs and novel drug targets, the decision surrounding proper therapy and drug dose, and the prediction of drug efficacy and adverse drug reactions. Results and Conclusion: In this article, we discussed the history, workflow, and advantages and disadvantages of machine learning techniques in the processes mentioned above. Although the advantages of machine learning techniques are fairly obvious, the application of machine learning techniques is currently limited. With further research, the application of machine techniques in drug development could be much more widespread and could potentially be one of the major methods used in drug development.

Background: Hepatitis C virus (HCV) infection is a global burden. There is no peptide vaccine found as modality to cure the disease is available due to the weak cellular immune response and the limitation to induce humoral immune response. Methods: Five predominated HCV subtypes in Indonesia (1a, 1b, 1c, 3a, and 3k) were aligned and the conserved regions were selected. Twenty alleles of class I MHC including HLA-A, HLA-B, and HLAC types were used to predict the potential epitopes by using NetMHCPan and IEDB. Eight alleles of HLA-DRB1, together with a combination of 3 alleles of HLA-DQA1 and 5 alleles of HLA-DQB1 were utilized for Class II MHC epitopes prediction using NetMHCIIPan and IEDB. LBtope and Ig- Pred were used to predict B cells epitopes. Moreover, proteasome analysis was performed by NetCTL and the stability of the epitopes in HLA was calculated using NetMHCStabPan for Class I. All predicted epitopes were analyzed for its antigenicity, toxicity, and stability. Population coverage, molecular docking and molecular dynamics were performed for several best epitopes. Results: The results showed that two best epitopes from envelop protein, GHRMAWDMMMNWSP (E1) and PALSTGLIHLHQN (E2) were selected as promising B cell and CD8+ T cell inducers. Other two peptides, LGIGTVLDQAETAG and VLVLNPSVAATLGF, taken from NS3 protein were selected as CD4+ T cell inducer. Conclusion: This study suggested the utilization of all four peptides to make a combinational peptide vaccine for in vivo study to prove its ability in inducing secondary response toward HCV.

Background: Paraoxonase 1 (PON1) is a paraoxonase, arylesterase and lactonase associated with protection of lipoproteins and cell membranes against oxidative modification. Objective: Based on antioxidative properties of PON1 and significance of 1,3,4-thiadiazoles in pharmaceutical chemistry, herein we aimed to evaluate the potentials of 1,3,4-thiadiazole derivatives as PON1 activators. Methods: 2-[[5-(2,4-Difluoro/dichlorophenylamino)-1,3,4-thiadiazol-2-yl]thio]acetophenone derivatives (1-18) were in vitro evaluated for their activator effects on PON1 which was purified using ammonium sulfate precipitation (60-80%) and DEAE-Sephadex anion exchange chromatography. Molecular docking studies were performed for the detection of affinities of all compounds to the active site of PON1. Moreover, Absorption, Distribution, Metabolism and Excretion (ADME) properties of all compounds were also in silico predicted. In silico molecular docking and ADME studies were carried out according to modules of Schrodinger’s Maestro molecular modeling package. Results: All compounds, particularly compounds 10, 13 and 17, were determined as promising PON1 activators and apart from compound 1, all of them were detected in the active site of PON1. Besides, ADME results indicated that all compounds were potential orally bioavailable drug-like molecules. Conclusion: PON1 activators, compounds 10, 13 and 17 stand out as potential drug candidates for further antioxidant studies and these compounds can be investigated for their therapeutic effects in many disorders such as atherosclerosis, diabetes mellitus, obesity, chronic liver inflammation and many more.

Background: Overexpression of Histone deacetylase 1 (HDAC1) is responsible for carcinogenesis by promoting epigenetic silence of tumour suppressor genes. Thus, HDAC1 inhibitors have emerged as the potential therapeutic leads against multiple human cancers, as they can block the activity of particular HDACs, renovate the expression of several tumour suppressor genes and bring about cell differentiation, cell cycle arrest and apoptosis. Methods: The present research work comprises atom-based 3D-QSAR, docking, molecular dynamic simulations and DFT (density functional theory) studies on a diverse series of hydroxamic acid derivatives as selective HDAC1 inhibitors. Two pharmacophoric models were generated and validated by calculating the enrichment factors with the help of the decoy set. The Four different 3D-QSAR models i.e., PLS (partial least square) model, MLR (multiple linear regression) model, Field-based model and GFA (Genetic function approximation) model were developed using ‘PHASE’ v3.4 (Schrödinger) and Discovery Studio (DS) 4.1 software and validated using different statistical parameters like internal and external validation. Results and Discussion: The results showed that the best PLS model has R2=0.991 and Q2=0.787, the best MLR model has R2= 0.993 and Q2= 0.893, the best Field-based model has R2= 0.974 and Q2= 0.782 and the best GFA model has R2= 0.868 and Q2= 0.782. Cross-validated coefficients, (rcv 2) of 0.967, 0.926, 0.966 and 0.829 was found for PLS model, MLR, Field based and GFA model, respectively, indicated the satisfactory correlativity and prediction. The docking studies were accomplished to find out the conformations of the molecules and their essential binding interactions with the target protein. The trustworthiness of the docking results was further confirmed by molecular dynamics (MD) simulations studies. Density Functional Theory (DFT) study was performed which promptly optimizes the geometry, stability and reactivity of the molecule during receptor-ligand interaction. Conclusion: Thus, the present research work provides spatial fingerprints which would be beneficial for the development of potent HDAC1 inhibitors.

Background: QSAR models as PLS, GFA, and 3D were developed for a series of matriptase inhibitors using 35 piperidyl-cyclohexylurea compounds. The training and test sets were divided into a set of 28 and 8 compounds, respectively and the pki values of each compound were used in the analysis. Methods: Docking and alignment methodologies were used to develop models in 3D QSAR. The best models among all were selected on the basis of regression statistics as r2, predictive r2 and Friedman Lack of fit measure. Hydrogen donors and rotatable bonds were found to be positively correlated properties for this target. The models were validated and used for the prediction of new compounds. Based on the predictions of 3D-QSAR model, 17 new compounds were prepared and their activities were predicted and compared with the active compound. Prediction of activities was performed for these 18 compounds using consensus results of all models. ADMET was also performed for the best-chosen compound and compared with the known active. Results and Conclusion: The developed model was able to validate the obtained results and can be successfully used to predict new potential and active compounds.

Background: The Quantitative structure activity relationship for thirty two novel substituted quinoxalines was performed for their antitubercular (Mycobacterium tuberculosis H37Rv) and antileptospiral (Leptospirainterrogans) activities. The quinoxalines were substituted with azetidinones, thiazolidinones and fluoroquinolones. Several compounds exhibited good activity against both the infections and they all possess fluoroquinolone moiety with the quinoxaline. Methods: The models developed showed good linear relationship (r2 = 0.71-0.88), with an internal predictive ability (q2> 0.61) and good external predictive ability (pred_r2>0.71). The compounds were separated into a training set on which regression was performed and a test set on which the predictive ability of the model was tested. Other statistical parameters including Ro2, Ro’2, k, k’ and Z- score were in the acceptable range. Results and Conclusion: The descriptors obtained explained the necessity of spatial orientation of atoms including branching and adjacency, presence of electronegative groups, balance between lipophilic elements and their binding strengths.