QSAR Data Research Articles

Patch seriation to visualize data and model parameters

We developed a new seriation merit function for enhancing the visual information of data matrices. A local similarity matrix is calculated, where the average similarity of neighbouring objects is calculated in a limited variable space and a global function is constructed to maximize the local similarities and cluster them into patches by simple row and column ordering. The method identifies data clusters in a powerful way, if the similarity of objects is caused by some variables and these variables differ for the distinct clusters. The method can be used in the presence of missing data and also on more than two-dimensional data arrays. We show the feasibility of the method on different data sets: on QSAR, chemical, material science, food science, cheminformatics and environmental data in two- and three-dimensional cases. The method can be used during the development and the interpretation of artificial neural network models by seriating different features of the models. It helps to identify interpretable models by elucidating clusters of objects, variables and hidden layer neurons.Graphical

The differential binding and biological efficacy of auxin herbicides.

Auxin herbicides have been used for selective weed control for 75 years and they continue to be amongst the most widely used weed control agents globally. The auxin herbicides fall into five chemical classes, with two herbicides not classified, and in all cases it is anticipated that recognition in the plant starts with binding to the Transport Inhibitor Response 1 (TIR1) family of auxin receptors. There is evidence that some classes of auxins act selectively with certain clades of receptors, although a comprehensive structure-activity relationship has not been available. Using purified receptor proteins to measure binding efficacy we have conducted quantitative structure activity relationship (qSAR) assays using representative members of the three receptor clades in Arabidopsis, TIR1, AFB2 and AFB5. Complementary qSAR data for biological efficacy at the whole-plant level using root growth inhibition and foliar phytotoxicity assays have also been analyzed for each family of auxin herbicides, including for the afb5-1 receptor mutant line. Comparisons of all these assays highlight differences in receptor selectivity and some systematic differences between results for binding in vitro and activity in vivo. The results could provide insights into weed spectrum differences between the different classes of auxin herbicides, as well as the potential resistance and cross-resistance implications for this herbicide class. © 2022 The Authors. Pest Management Science published by John Wiley & Sons Ltd on behalf of Society of Chemical Industry.

Ridge regression coupled with a new uninformative variable elimination algorithm as a new descriptor screening method: Application of data reduction in QSAR study of some sulfonated derivatives as c-Met inhibitors

An uninformative variable elimination algorithm was combined with the Ridge regression method. This combination makes the penalized Ridge method, which is essentially incapable of deleting or selecting a variable, an efficient descriptor screening method for screening of descriptors in QSAR studies. The importance of descriptors was used to identify uninformative and redundant descriptors. The descriptor importance was defined as the ratio of the mean coefficients to the standard deviation of the coefficients derived from the leave-one-out (LOO) Ridge models. Then, with an iterative algorithm, among 392 molecular descriptors obtained after initial preprocessing, 358 uninformative descriptors were identified and removed from the data set. The remaining 34 descriptors were used for further variable selection and modeling through the smoothly clipped absolute deviation (SCAD) method. The proposed screening method was used to screen and reduce the size of QSAR data containing 63 compounds as c-Met inhibitors with anti-cancer activity. For comparison, the SCAD models were constructed using original high-dimensional data (SCAD), and the reduced-size data via the UVE-Ridge screening method (UVE-Ridge-SCAD). Both models were used to predict biological activities (pIC50) for compounds in the external test set. The coefficients of determination (R2) and the FIT parameter for the test data predicted using the SCAD model (with 12 selected descriptors) were equal to 0.8042 and 3.78, respectively. The UVE-Ridge-SCAD model (with 9 descriptors) showed R2 and FIT values of 0.9025 and 4.38, respectively. Higher R2 and FIT values indicate the excellent predictive power of the UVE-Ridge-SCAD model. Finally, the UVE-Ridge-SCAD model was used to suggest new high-potency compounds. The accuracy of the proposed compounds was investigated using ligand-receptor interactions.

Microbiological Aspects of Unique, Rare, and Unusual Fatty Acids Derived from Natural Amides and Their Pharmacological Profile

In the proposed review, the pharmacological profile of unique, rare, and unusual fatty acids derived from natural amides is considered. These amides are produced by various microorganisms, lichens, and fungi. The biological activity of some natural fatty acid amides has been determined by their isolation from natural sources, but the biological activity of fatty acids has not been practically studied. According to QSAR data, the biological activity of fatty acids is shown, which demonstrated strong antifungal, antibacterial, antiviral, antineoplastic, anti-inflammatory activities. Moreover, some fatty acids have shown rare activities such as antidiabetic, anti-infective, anti-eczematic, antimutagenic, and anti-psoriatic activities. For some fatty acids that have pronounced biological properties, 3D graphs are shown that show a graphical representation of unique activities. These data are undoubtedly of both theoretical and practical interest for chemists, pharmacologists, as well as for the pharmaceutical industry, which is engaged in the synthesis of biologically active drugs.

Compatibility of Diborane and Borane with FOX-7 - A DFT Treatment

FOX-7 as an energetic material is getting more and more popular as the constituents of various ammunitions. On the other hand, boronic species attract attention as fuels in rocket engineering. The present study, within the constraints of density functional theory, considers some composites of them, that is FOX-7+B2H6, FOX-7+2BH3 and FOX-7+BH3. The calculations at the B3LYP/6-311++G(d,p) level indicate that all the components are structurally stable in the composites although they interact with each other electronically. Various quantum chemical and QSAR data are obtained and discussed.

Some Curcumin Isomers and Their Enol Tautomers - A DFT Treatment

Curcumin is a well known natural product having some health benefits. In the present study, within the constraints of density functional theory (at the level of B3LYP/6-31G(d,p)), some configurational isomers of curcumin and their keto-enol tautomers have been investigated. Some quantum chemical, QSAR and spectral data of them have been obtained and discussed.

Recent Trends in Drug Design and Discovery.

Structure-based drug design is a wide area of identification of selective inhibitors of a target of interest. From the time of the availability of three dimensional structure of the drug targets, mostly the proteins, many computational methods had emerged to address the challenges associated with drug design process. Particularly, drug-likeness, druggability of the target protein, specificity, off-target binding, etc., are the important factors to determine the efficacy of new chemical inhibitors. The aim of the present research was to improve the drug design strategies in field of design of novel inhibitors with respect to specific target protein in disease pathology. Recent statistical machine learning methods applied for structural and chemical data analysis had been elaborated in current drug design field. As the size of the biological data shows a continuous growth, new computational algorithms and analytical methods are being developed with different objectives. It covers a wide area, from protein structure prediction to drug toxicity prediction. Moreover, the computational methods are available to analyze the structural data of varying types and sizes of which, most of the semi-empirical force field and quantum mechanics based molecular modeling methods showed a proven accuracy towards analysing small structural data sets while statistics based methods such as machine learning, QSAR and other specific data analytics methods are robust for large scale data analysis. In this present study, the background has been reviewed for new drug lead development with respect specific drug targets of interest. Overall approach of both the extreme methods were also used to demonstrate with the plausible outcome. In this chapter, we focus on the recent developments in the structure-based drug design using advanced molecular modeling techniques in conjunction with machine learning and other data analytics methods. Natural products based drug discovery is also discussed.

Effect of nano metal oxides on heme molecule: molecular and biomolecular approaches

Interaction of components of living cells with various nanomaterials in the gas phase has been one of extensive concern since they become intensively utilized in various life aspects. This work is carried out to investigate the interaction between heme molecule, as the main component of hemoglobin, with several familiar and non-familiar divalent structures such as O2, CO2, CO, MgO, CoO, NiO, CuO and ZnO. Geometry optimization processes as well as QSAR descriptors are conducted using semiemprical quantum mechanical calculations at PM6 level. Results illustrate that adsorbing O2 and CO on heme lowers their TDM helping heme in performing its transportation function and not interacting with other species. On the other hand, when CoO and ZnO interacting with heme the TDM of the resultant structures increase greatly reflecting high reactivity which may interact with other species more than performing its function. Therefore, interacting species other than O2 may disturb the transportation function of heme structure. QSAR data of IP regarding interaction of O2 with heme ensure the TDM result that reflects lowering its activity. IP of H-CO adsorbed is the lowest indicating high reactivity while those of H-O2, H-CO2, H-MgO and H-NiO in the complex form are the highest values indicating that it is difficult to form a complex structure with them. Therefore, heme interactions with structures rather than O2 and CO2 affects negatively its function as gas transporter.

Green synthesis approach for novel benzenesulfonamide nanometer complexes with elaborated spectral, theoretical and biological treatments

New series of nano‐sized bi‐homonuclear Ce (III), ZrO (II), Sn (II), Pb (II), Cr (III), Fe (III) and Cu (II) complexes with 4‐[(2,4‐dihydroxybenzylidene)amino]‐N‐(1,3‐thiazol‐2‐yl) benzenesulfonamide (H3L) were synthesized via green solid‐state method. The structural and molecular formulae of all synthesized complexes were established based on variable spectral, analytical and theoretical implementations. FT‐IR study confirms the coordination of H3L with metal ions through the Schiff base and sulfonamide centers in di‐basic tetra‐dentate mode. Thermal analysis, magnetic moment and electronic spectra are attributing to octahedral configuration around Ce (III), Cr (III) and Fe (III) centers, while with ZrO (II), Sn (II), Pb (II) and Cu (II) centers, acquired tetrahedral arrangement. TEM and XRD studies, represent the nanometer characters of most metal ion complexes. TGA curves are utilized to compute the activation thermo‐kinetic parameters over different decomposition stages applying Coats‐Redfern method. Theoretical implementation executed by Gaussian09 program exerted the structures for the best atomic orientation over whole molecules. QSAR data were achieved over Hyper Chem 8.1 program through molecular mechanics process. Docking complexes between free ligand and different protein receptors were obtained through AutoDock Tools 4.2. Antimicrobial, antifungal and antitumor activities of the metal complexes were studied in comparing with free ligand to assert their potential therapeutic uses. H3L, Ce (III), Fe (III) and Cu (II) complexes displayed high antibacterial activity near that of standard Gentamycin. Moreover, Cr (III) complex displayed highest cytotoxicity against human liver Carcinoma cell line (HEPG2).

QSAR, docking studies and toxicology prediction of isoquinoline derivatives as leucine aminopeptidase inhibitors

Here, we report a theoretical study on leucine aminopeptidase (LAP) inhibitors. 3D quantitative structure–activity relationship (3D-QSAR) studies were carried out on a series of known substituted dihydroisoquinoline derivatives and other chemical compounds with inhibitory activity towards LAP. We also designed a series of 33 isoquinoline derivatives containing phosphonic/phosphinic acid moieties and applied the 3D-QSAR model to evaluate their possible LAP inhibitory activity. In silico studies showed that these compounds interact with zinc ions, form H-bonds and exhibit hydrophobic interactions with amino acids in the LAP binding pocket as well as present high scoring values. The obtained QSAR data provide useful information about the structural characteristics of inhibitors which contribute to their inhibitory potency.

Tales from the war on error: the art and science of curating QSAR data.

Curating the data underlying quantitative structure-activity relationship models is a never-ending struggle. Some curation can now be automated but much cannot, especially where data as complex as those pertaining to molecular absorption, distribution, metabolism, excretion, and toxicity are concerned (vide infra). The authors discuss some particularly challenging problem areas in terms of specific examples involving experimental context, incompleteness of data, confusion of units, problematic nomenclature, tautomerism, and misapplication of automated structure recognition tools.

The Permeability of an Artificial Membrane for Wide Range of pH in Human Gastrointestinal Tract: Experimental Measurements and Quantitative StructureActivity Relationship.

In silico models for membrane permeability have been based on values measured for single pH. Depending on the diet (fasted/fed state) and part of human intestine the range of pH varies approximately from 2.4 to 8.0. This motivated to study and model the membrane permeability of chemicals considering the whole range of pH in the human intestine. For this, effective membrane permeability values were measured for 65 drugs and drug-like compounds using PAMPA method at four pHs (3, 5, 7.4, 9) over 48 h, introducing technological innovations for the time-dependence measurement. The highest permeability value of a compound from four pHs was used to derive QSAR analyzing a large pool of molecular descriptors and introducing new descriptor. Using stepwise forward selection approach a significant QSAR model was derived that included only two mechanistically relevant descriptors, the logarithmic octanol-water partition coefficient and hydrogen bonding surface area. Prediction confidence of the model was blind tested with a true external validation set of 15 compounds. The resulting QSAR model shows potential to combine permeability values from various pH-s into one descriptive and predictive model for estimating maximum permeability in human gastrointestinal tract. The QSAR model and data are available through the QsarDB repository (http://dx.doi.org/10.15152/QDB.137).

Expanding the horizons of G protein-coupled receptor structure-based ligand discovery and optimization using homology models.

With >800 members in humans, the G protein-coupled receptor (GPCR) super-family is the target for more than 30% of the marketed drugs. The recent boom in GPCR crystallography has enabled the solution of ∼30 different GPCR structures, which boosted the identification and optimization of novel modulators and new chemical entities through structure-based strategies. However, the number of available structures represents a small part of the human GPCR druggable target space, and its complete coverage in the near future seems unlikely. Homology modelling represents a reliable tool to fill this gap, and hence to vastly expand the horizons of structure-based drug discovery and design. In this Feature Article, we show from a wealth of retrospective and prospective studies that in spite of the pitfalls of and standing challenges in homology modelling, structural models have been critical for the blossoming and success of GPCR structure-based lead discovery and optimization endeavours; in addition, they have also been instrumental in characterizing receptor-ligand interaction, guiding the design of site-directed mutagenesis and SAR studies, and assessing off-target effects. Considering though their current limitations, we also discuss the most pressing issues to develop more accurate homology modelling strategies, with a special focus on the integration of computational tools with biochemical, biophysical and QSAR data, highlighting methodological aspects and recent progress. The teachings of the three GPCR Dock community-wide assessments and the fresh developments in GPCR classes B, C and F are commented. This is a fast growing and highly promising field of research, and in the coming years, the use of high-quality models should enable the discovery of a growing number of potent, selective and efficient GPCR drug leads with high therapeutic potential through receptor structure-based strategies.

Benchmarking study of parameter variation when using signature fingerprints together with support vector machines.

QSAR modeling using molecular signatures and support vector machines with a radial basis function is increasingly used for virtual screening in the drug discovery field. This method has three free parameters: C, γ, and signature height. C is a penalty parameter that limits overfitting, γ controls the width of the radial basis function kernel, and the signature height determines how much of the molecule is described by each atom signature. Determination of optimal values for these parameters is time-consuming. Good default values could therefore save considerable computational cost. The goal of this project was to investigate whether such default values could be found by using seven public QSAR data sets spanning a wide range of end points and using both a bit version and a count version of the molecular signatures. On the basis of the experiments performed, we recommend a parameter set of heights 0 to 2 for the count version of the signature fingerprints and heights 0 to 3 for the bit version. These are in combination with a support vector machine using C in the range of 1 to 100 and γ in the range of 0.001 to 0.1. When data sets are small or longer run times are not a problem, then there is reason to consider the addition of height 3 to the count fingerprint and a wider grid search. However, marked improvements should not be expected.

Histamine H4 receptor ligands: future applications and state of art.

Histamine is a chemical transmitter found practically in whole organism and exerts its effects through the interaction with H1 to H4 histaminergic receptors. Specifically, H4 receptors are found mainly in immune cells and blood-forming tissues, thus are involved in inflammatory and immune processes, as well as some actions in central nervous system. Therefore, H4 receptor ligands can have applications in the treatment of chronic inflammatory and immune diseases and may be novel therapeutic option in these conditions. Several H4 receptor ligands have been described from early 2000's until nowadays, being imidazole, indolecarboxamide, 2-aminopyrimidine, quinazoline, and quinoxaline scaffolds the most explored and discussed in this review. Moreover, several studies of molecular modeling using homology models of H4 receptor and QSAR data of the ligands are summarized. The increasing and promising therapeutic applications are leading these compounds to clinical trials, which probably will be part of the next generation of blockbuster drugs.

Assessment of Machine Learning Reliability Methods for Quantifying the Applicability Domain of QSAR Regression Models

The vastness of chemical space and the relatively small coverage by experimental data recording molecular properties require us to identify subspaces, or domains, for which we can confidently apply QSAR models. The prediction of QSAR models in these domains is reliable, and potential subsequent investigations of such compounds would find that the predictions closely match the experimental values. Standard approaches in QSAR assume that predictions are more reliable for compounds that are "similar" to those in subspaces with denser experimental data. Here, we report on a study of an alternative set of techniques recently proposed in the machine learning community. These methods quantify prediction confidence through estimation of the prediction error at the point of interest. Our study includes 20 public QSAR data sets with continuous response and assesses the quality of 10 reliability scoring methods by observing their correlation with prediction error. We show that these new alternative approaches can outperform standard reliability scores that rely only on similarity to compounds in the training set. The results also indicate that the quality of reliability scoring methods is sensitive to data set characteristics and to the regression method used in QSAR. We demonstrate that at the cost of increased computational complexity these dependencies can be leveraged by integration of scores from various reliability estimation approaches. The reliability estimation techniques described in this paper have been implemented in an open source add-on package ( https://bitbucket.org/biolab/orange-reliability ) to the Orange data mining suite.

Radial basis function network-based transformation for nonlinear partial least-squares as optimized by particle swarm optimization: Application to QSAR studies

This study presented a new version of the nonlinear partial least-squares method on the optimized radial basis function network transformation by particle swarm optimization (PSORBFPLS). This algorithm firstly transformed the training inputs into the hidden outputs using the nonlinear transformation carried by a radial basis function network (RBFN), and then employed linear partial least-squares (PLS) to relate the outputs of the hidden layer to the bioactivities. The widths and centers involved in RBF transformation were optimized by particle swarm optimization (PSO) with the minimized model error via PLS modeling as the criterion. The number of latent variables associated with PLS modeling was automatically identified by F-statistic. Two QSAR data sets were used to evaluate the performance of the newly proposed PSORBFPLS. Results of these two data sets demonstrated that PSORBFPLS offers substantially enhanced capacities in modeling nonlinearity while circumvents overfitting frequently encountered in nonlinear modeling.

Unsupervised selection of informative descriptors in QSAR study of anti-HIV activities of HEPT derivatives

With the increasing ease of measuring and calculating multiple descriptors per molecule in quantitative structure–activity relationship, the importance of variable selection for data reduction and improving interpretability is gaining importance. While variable selection has been extensively studied in the context of supervised learning, in this paper, an unsupervised learning method is proposed for variable selection and its performance is assessed using a typical QSAR data set. Whereas there is no real dependent variable in the proposed variable selection algorithm, applied variable selection is unsupervised indeed. Besides, scores that are the linear combination of the data variables are set as dependent variables (artificial dependent variables). It includes 107 derivatives of HEPT molecule, characterized by 160 descriptors encoding the steric, hydrophobic, electronic and structural features of HEPT derivatives. The aims of this procedure are generating a subset of descriptors from a data set with the relevant variables, eliminating redundancy, and reducing multicollinearity. The core of this methodology is based on jack-knife resampling method. In this paper, using jack-knife led to selection of 48 out of 160 initial descriptors, so that the data information was preserved. Lastly, using influence effect on prediction resulted in eight descriptors as representative of the 160 descriptors. Constructed model with final 8 descriptors has Q2IN=0.67, R2=0.74, Q2EXT=0.85. It represents adequacy of our strategy for preserving data structure.

Ligand Based Validated Comparative Chemometric Modeling and Pharmacophore Mapping of Aurone Derivatives as Antimalarial Agents

Chloroquine resistance is nowadays a great problem. Aurone derivatives are effective against chloroquine resistant parasite. Ligand based validated comparative chemometric modeling through 2D-QSAR and kNN-MFA 3DQSAR studies as well as common feature 3D pharmacophore mapping were done on thirtyfive aurone derivatives having antimalarial activity. Statistically significant 2D-QSAR models were generated on unsplitted as well as splitted dataset by MLR and PLS technique. The MLR model of the unsplitted method was validated by two-deep cross validation and 10 fold cross validation for determining the predictive ability. The PLS technique of the unsplitted method was done to compare the significance of these methods. In the splitted method, model was developed on the training set by Y-based ranking method by using the same descriptors and was validated on fifty pairs of the test and the training sets by k-MCA technique. These models generated by using the same descriptors were well validated irrespective of MLR as well as PLS analysis of unsplitted as well as splitted methods and are showing similar results. Therefore, these descriptors and model generated were reliable and robust. The kNN-MFA 3D-QSAR models were generated by three variable selection methods: genetic algorithm, simulated annealing and stepwise regression. The kNN-MFA 3D-QSAR results support the 2D QSAR data and in turn validate the earlier observed SAR results. Common feature 3D-pharmacophore generation was performed on these compounds to validate both 2D and 3D-QSAR studies as well as the earlier observed SAR data. The work highlights the required structural features for the higher antimalarial activity.

Substrate versus inhibitor dynamics of P‐glycoprotein

By far the most studied multidrug resistance protein is P-glycoprotein. Despite recent structural data, key questions about its function remain. P-glycoprotein (P-gp) is flexible and undergoes large conformational changes as part of its function and in this respect, details not only of the export cycle, but also the recognition stage are currently lacking. Given the flexibility, molecular dynamics (MD) simulations provide an ideal tool to examine this aspect in detail. We have performed MD simulations to examine the behaviour of P-gp. In agreement with previous reports, we found that P-gp undergoes large conformational changes which tended to result in the nucleotide-binding domains coming closer together. In all simulations, the approach of the NBDs was asymmetrical in agreement with previous observations for other ABC transporter proteins. To validate the simulations, we make extensive comparison to previous cross-linking data. Our results show very good agreement with the available data. We then went on to compare the influence of inhibitor compounds bound with simulations of a substrate (daunorubicin) bound. Our results suggest that inhibitors may work by keeping the NBDs apart, thus preventing ATP-hydrolysis. On the other hand, repeat simulations of daunorubicin (substrate) in one particular binding pose suggest that the approach of the NBDs is not impaired and that the structure would be still be competent to perform ATP hydrolysis, thus providing a model for inhibition or substrate transport. Finally we compare the latter to earlier QSAR data to provide a model for the first part of substrate transport within P-gp.