QSAR studies of TIBO derivatives as HIV-1 reverse transcriptase inhibitors using HQSAR, CoMFA and CoMSIA

Comparison of various methods for validity evaluation of QSAR models

BACKGROUND

Quantitative structure-activity relationship (QSAR) modeling is one of the most important computational tools employed in drug discovery and development. The external validation of QSAR models is the main point to check the reliability of developed models for the prediction activity of not yet synthesized compounds. It was performed by different criteria in the literature.

METHODS

In this study, 44 reported QSAR models for biologically active compounds reported in scientific papers were collected. Various statistical parameters of external validation of a QSAR model were calculated, and the results were discussed.

RESULTS

The findings revealed that employing the coefficient of determination (r²) alone could not indicate the validity of a QSAR model. The established criteria for external validation have some advantages and disadvantages which should be considered in QSAR studies.

CONCLUSION

This study showed that these methods alone are not only enough to indicate the validity/invalidity of a QSAR model.

Research on the anti-tumor activity of a novel aminopeptidase inhibitor based on 3D QSAR model

Background:

Aminopeptidase N (APN) is a type II transmembrane zinc ion-dependent metalloprotease. It is closely related to many processes of tumor occurrence and development, such as the formation of new blood vessels and tumor metastasis. Recent studies have shown that APN is a member of the family of surface markers of liver cancer stem cells. Therefore, APN small molecule inhibitors may have multiple compound functions, exerting multiple anti-tumor effects at multiple stages of cancer occurrence and development.

Methods:

Comparative Molecular Field Analysis (CoMFA) and Comparative Molecular Similarity Indices Analysis (CoMSIA) approaches.

Results:

Both internal and external cross-validation were conducted to obtain high predictive and satisfactory CoMFA model (q2 = 0.627, r2 =0.995, SEE = 0.043) and CoMSIA model (q2 = 0.575, r2 = 0.998, SEE = 0.031). The statistical results obtained from CoMFA and CoMSIA models were full of credibility and remarkable predictive power.

Conclusion:

The results of 3D-QSAR are reliable and significant with high predictive (q2) ability, and a lower value of the standard error of estimation indicates a good correlation between predicted and observed activity. All these results revealed many useful structural insights to improve the activity of the newly designed APN small molecule inhibitors.

Topomer CoMFA and HQSAR Study on Benzimidazole Derivative as NS5B Polymerase Inhibitor

Background:

In recent years, the number of people infected with the hepatitis C virus (HCV) is increasing rapidly. This has become a major threat to global health, therefore, new anti- HCV drugs are urgently needed. HCV NS5B polymerase is an RNA-dependent RNA polymerase (RdRp), which plays an important role in virus replication, and can effectively prevent the replication of HCV sub-genomic RNA in daughter cells. It is considered a very promising HCV therapeutic target for the design of anti-HCV drugs.

Methods:

In order to explore the relationship between the structure of benzimidazole derivative and its inhibitory activity on NS5B polymerase, holographic quantitative structure-activity relationship (HQSAR) and Topomer comparative molecular field analysis (CoMFA) were used to establish benzimidazole QSAR model of derivative inhibitors.

Results:

The results show that for the Topomer CoMFA model, the cross-validation coefficient q2 value is 0.883, and the non-cross-validation coefficient r2 value is 0.975. The model is reasonable, reliable, and has a good predictive ability. For the HQSAR model, the cross-validated q2 value is 0.922, and the uncross-validated r2 value is 0.971, indicating that the model data fit well and has a high predictive ability. Through the analysis of the contour map and color code diagram, 40 new benzimidazole inhibitor molecules were designed, and all of them have higher activity than template molecules, and the new molecules have significant interaction sites with protein 3SKE.

Conclusion:

The 3D-QSAR model established by Topomer CoMFA and HQSAR has good prediction results and the statistical verification is valid. The newly designed molecules and docking results provide theoretical guidance for the synthesis of new NS5B polymerase inhibitors and for the identification of key residues that the inhibitors bind to NS5B, which helps to better understand their inhibitory mechanism. These findings are helpful for the development of new anti-HCV drugs.

Machine Learning-Based Scoring Functions, Development and Applications with SAnDReS

BACKGROUND

Analysis of atomic coordinates of protein-ligand complexes can provide three-dimensional data to generate computational models to evaluate binding affinity and thermodynamic state functions. Application of machine learning techniques can create models to assess protein-ligand potential energy and binding affinity. These methods show superior predictive performance when compared with classical scoring functions available in docking programs.

OBJECTIVE

Our purpose here is to review the development and application of the program SAnDReS. We describe the creation of machine learning models to assess the binding affinity of protein-ligand complexes.

METHODS

SAnDReS implements machine learning methods available in the scikit-learn library. This program is available for download at https://github.com/azevedolab/sandres. SAnDReS uses crystallographic structures, binding and thermodynamic data to create targeted scoring functions.

RESULTS

Recent applications of the program SAnDReS to drug targets such as Coagulation factor Xa, cyclin-dependent kinases and HIV-1 protease were able to create targeted scoring functions to predict inhibition of these proteins. These targeted models outperform classical scoring functions.

CONCLUSION

Here, we reviewed the development of machine learning scoring functions to predict binding affinity through the application of the program SAnDReS. Our studies show the superior predictive performance of the SAnDReS-developed models when compared with classical scoring functions available in the programs such as AutoDock4, Molegro Virtual Docker and AutoDock Vina.

Structural modification of 4, 5-dihydro-[1, 2, 4] triazolo [4, 3-f] pteridine derivatives as BRD4 inhibitors using 2D/3D-QSAR and molecular docking analysis

Cancer treatment continues to be one of the most serious public health issues in the world. The overexpression of BRD4 protein has led to a series of malignant tumors, hence the development of small molecule BRD4 protease inhibitors has always been a hot spot in the field of medical research. In this study, a series of 4,5-dihydro-[1, 2, 4] triazolo [4, 3-f] pteridine derivatives were used to establish 3D/2D-QSAR models and to discuss the relationship between inhibitor structure and activity. Four ideal models were established, including the comparative molecular field analysis (CoMFA: [Formula: see text] = 0.574, [Formula: see text] = 0.947) model, comparative molecular similarity index analysis (CoMSIA: [Formula: see text]= 0.622, [Formula: see text] = 0.916) model, topomer CoMFA ([Formula: see text] = 0.691, [Formula: see text]= 0.912) model and hologram quantitative structure-activity relationship (HQSAR: [Formula: see text]= 0.759, [Formula: see text] = 0.963) model. They show quite good external predictive power for the test set, with [Formula: see text] values of 0.602, 0.624, 0.671 and 0.750, respectively. In addition, the contour and color code map given by the 2D/3D-QSAR model with the results of molecular docking analyzed to chalk up modification methods for improving inhibitory activity, which was verified by designing novel compounds. The analysis results are helpful to promote the modification of the inhibitor framework and to provide a reference for the construction of new and promising BRD4 inhibitor compounds.

Combined HQSAR method and molecular docking study on genotoxicity mechanism of quinolones with higher genotoxicity

Using the hologram quantitative structure-activity relationship (HQSAR) method, a quantitative model of the structure-activity relationship between the genotoxicity of quinolones towards gram-negative bacteria and structure of quinolones is constructed. A series of novel quinolones are designed, and 4 environmentally friendly quinolone derivatives are finally selected, because of their enhanced genotoxicity towards gram-negative/positive bacteria, decreased bioconcentration and increased photodegradability and biodegradability. The mechanisms underlying the genotoxicity of quinolones and its derivatives are analysed based on amino acid residues and molecular interactions. Three hydrophilic amino acids [arginine (ARG), asparagine (ASN) and aspartic acid (ASP)] play important roles in the antibacterial effects of quinolones. The introduction of highly hydrophilic groups into the C-7 position of amifloxacin (AMI) not only improved the stability of the AMI derivative-topoisomerase IV-DNA complex but also improved the antibacterial activities of AMI derivatives.

Machine Learning to Predict Binding Affinity

Recent progress in the development of scientific libraries with machine-learning techniques paved the way for the implementation of integrated computational tools to predict ligand-binding affinity. The prediction of binding affinity uses the atomic coordinates of protein-ligand complexes. These new computational tools made application of a broad spectrum of machine-learning techniques to study protein-ligand interactions possible. The essential aspect of these machine-learning approaches is to train a new computational model by using technologies such as supervised machine-learning techniques, convolutional neural network, and random forest to mention the most commonly applied methods. In this chapter, we focus on supervised machine-learning techniques and their applications in the development of protein-targeted scoring functions for the prediction of binding affinity. We discuss the development of the program SAnDReS and its application to the creation of machine-learning models to predict inhibition of cyclin-dependent kinase and HIV-1 protease. Moreover, we describe the scoring function space, and how to use it to explain the development of targeted scoring functions.