Multi-way PLS modeling of structure-activity data by incorporating electrostatic and lipophilic potentials on molecular surface

We devised and elaborated a surface-based three-dimensional-quantitative structure-activity relationship (3D-QSAR) method, which had been proposed in the previous study. This approach can be applied to more general case where both the electrostatic and lipophilic potentials on molecular surface simultaneously change. The 3D coordinates of all sampling points on molecular surface are projected into a 2D map by Kohonen neural network (KNN). Each node in the map is coded by the associated molecular electrostatic potential (MEP) or molecular lipophilic potential (MLP) values. The electrostatic and lipophilic KNN maps are generated for each compound and the four-way array is constructed by collecting two KNN maps of all samples. The correlation between four-way array and biological activity is examined by four-way partial least-squares (PLS). For validation, the structure-activity data of estrogen receptor antagonists was investigated. The four-way PLS model gave the high statistics at calibration and validation stages. The coefficients of the four-way PLS model back-projected on molecular surface had a reasonable 3D distribution and it was nicely consistent with active site of the estrogen receptor which was recently made clear by X-ray crystallography.     

Generative topographic mapping of binding pocket of β2 receptor and three‐way partial least squares modeling of inhibitory activities

The visualization and characterization of protein pockets is the starting point for many structure‐based drug design projects. The size and shape of protein pockets dictate 3D geometry of ligands that can strongly inhibit the following biological events. Thus, a minimal requirement for inhibition is that a molecule sterically binds the active site with some allowance for induced fit. Methods for direct display of active sites in a protein have become prevalent in recent years. In this study, a new mapping method, generative topographic mapping, is investigated to describe the 3D surface of protein pocket. The β2 receptor protein is used as a benchmark. By mapping the molecular surface points and assigning the associated molecular electrostatic potential (MEP) values, the original 3D structure of the active site is well reproduced by the 2D latent map in generative topographic mapping. The distributions of MEP values of two 2D latent maps derived from the inhibitor and the β2 receptor protein are well complemented. Using three‐way partial least squares modeling, a predictive model linking the inhibitory activity and their MEP values can be constructed, which was not feasible in the previous spherical self‐organizing map studies. The resulting regression coefficient matrix of the three‐way partial least squares model has many insights for understanding the structural requirements for β2 inhibitory activity. Copyright © 2014 John Wiley & Sons, Ltd.     

Representation of molecular electrostatic potentials of biopolymer by self‐organizing feature map

The Kohonen self‐organizing map was introduced to map the protein molecular surface features. The protein or polypeptide properties, such as shape and molecular electrostatic potential, can be visualized by self‐organizing map, which was trained by the 3D surface coordinates. Such maps allow the visual comparison of molecular properties between proteins having common topological or chemical features.     

The comparison of geometric and electronic properties of molecular surfaces by neural networks: Application to the analysis of corticosteroid-binding globulin activity of steroids

Summary It is shown how a self-organizing neural network such as the one introduced by Kohonen can be used to analyze features of molecular surfaces, such as shape and the molecular electrostatic potential. On the one hand, two-dimensional maps of molecular surface properties can be generated and used for the comparison of a set of molecules. On the other hand, the surface geometry of one molecule can be stored in a network and this network can be used as a template for the analysis of the shape of various other molecules. The application of these techniques to a series of steroids exhibiting a range of binding activities to the corticosteroid-binding globulin receptor allows one to pinpoint the essential features necessary for biological activity.     

Regression with multiple regressor arrays

Extension of standard regression to the case of multiple regressor arrays is given via the Kronecker product. The method is illustrated using ordinary least squares regression (OLS) as well as the latent variable (LV) methods principal component regression (PCR) and partial least squares regression (PLS). Denoting the method applied to PLS as mrPLS, the latter was shown to explain as much or more variance for the first LV relative to the comparable L‐partial least squares regression (L‐PLS) model. The same relationship holds when mrPLS is compared to PLS or n‐way partial least squares (N‐PLS) and the response array is 2‐way or 3‐way, respectively, where the regressor array corresponding to the first mode of the response array is 2‐way and the second mode regressor array is an identity matrix. In a comparison with N‐PLS using fragrance data, mrPLS proved superior in a validation sense when model selection was used. Though the focus is on 2‐way regressor arrays, the method can be applied to n‐way regressors via N‐PLS. Copyright © 2007 John Wiley & Sons, Ltd.     

Molecular modeling, theoretical calculations and property evaluation of three muscarinic agonists. X-ray structures of LU 25-109 and WAL 2014

LU 25-109 (II) and WAL 2014 (talsaclidine, III) are two M1 muscarinic agonists chemically related to the natural substance arecoline (I). All these compounds have beneficial effects on memory and cognition in animals and humans, and they have been proposed in the treatment of Alzheimer's disease, but only III will likely find a place in therapy. In this work we have investigated the solid state structures of II and III, and the X-ray structures of the two molecules and of the parent compound I have been used to input a series of computational chemistry efforts.

In particular, the X-ray geometries have been manipulated to model 20 molecular structures (1–20) which have been submitted to ab initio, semiempirical quantum mechanics and molecular mechanics calculations. The conformational space accessible to the 20 structures has been assessed by means of potential energy maps. The reactivities of 1–20 have been estimated by examining at the graphics terminal the composition and the extension of the frontier orbitals (HOMOs and LUMOs) and of the molecular electrostatic potential. The information obtained has been interpreted to explain the different degrees of activity shown by I–III. Our data indicate that III has better in vivo activity for its intermediate size, less polar surface, conformational rigidity and orientation of reactive domains.     

Multiconformational composite molecular potential fields in the analysis of drug action. I. Methodology and first evaluation using 5-HT and histamine action as examples

Summary The quality of molecular electrostatic maps generated by non-quantum mechanical methods has been improved using extended electron distributions. Further simplification has been achieved by distilling these maps down to their energy extrema. A new means of defining surface interaction has been added and the resulting composite map has been plotted for a limited number of low-lying conformers of a series of agonists and antagonists of the H₂ and H₃ receptors and 5-HT_1A and 5-HT_1D receptors. The results from the cross-comparison of these maps indicate their ability to distinguish the specific receptor. Interesting consequences of the method are that structural overlay is irrelevant, that several conformations may contribute to the overall binding pattern and that lesser pharmacological activities may be deduced from the results.     

Molecular modeling of the intestinal bile acid carrier: A comparative molecular field analysis study

A structure–binding activity relationship for the intestinal bile acidtransporter has been developed using data from a series of bile acid analogsin a comparative molecular field analysis (CoMFA). The studied compoundsconsisted of a series of bile acid–peptide conjugates, withmodifications at the 24 position of the cholic acid sterol nucleus, andcompounds with slight modifications at the 3, 7, and 12 positions. For theCoMFA study, these compounds were divided into a training set and a test set,comprising 25 and 5 molecules, respectively. The best three-dimensionalquantitative structure–activity relationship model found rationalizesthe steric and electrostatic factors which modulate affinity to the bile acidcarrier with a cross-validated, conventional and predictive r²of 0.63, 0.96, and 0.69, respectively, indicating a good predictive model forcarrier affinity. Binding is facilitated by positioning an electronegativemoiety at the 24–27 position, and also by steric bulk at the end of theside chain. The model suggests substitutions at positions 3, 7, 12, and 24that could lead to new substrates with reasonable affinity for the carrier.     

An automated method for predicting the positions of hydrogen-bonding atoms in binding sites

Hydrogen bonds are the most specific, and therefore predictable of the intermolecular interactions involved in ligand–protein binding. Given the structure of a molecule, it is possible to estimate the positions at which complementary hydrogen-bonding atoms could be found. Crystal-survey data are used in the design of a program, HBMAP, that generates a hydrogen-bond map for any given ligand, which contains all the feasible positions at which a complementary atom could be found. On superposition of ligands, the overlapping regions of their maps represent positions of receptor atoms to which each molecule can bind. The certainty of these positions is increased by the incorporation of a larger number and diversity of molecules. In this work, superposition is achieved using the program HBMATCH, which uses simulated annealing to generate the correspondence between points from the hydrogen-bonding maps of the two molecules. Equivalent matches are distinguished on the basis of their steric similarity. The strategy is tested on a number of ligands for which ligand–protein complexes have been solved crystallographically, which allows validation of the techniques. The receptor atom positions of thermolysin are successfully predicted when the correct superposition is obtained.     

Heuristic lipophilicity potential for computer-aided rational drug design

In this contribution we suggest a heuristic molecular lipophilicitypotential (HMLP), which is a structure-based technique requiring noempirical indices of atomic lipophilicity. The input data used in thisapproach are molecular geometries and molecular surfaces. The HMLP is amodified electrostatic potential, combined with the averaged influences fromthe molecular environment. Quantum mechanics is used to calculate theelectron density function (r) and the electrostatic potential V(r), andfrom this information a lipophilicity potential L(r) is generated. The HMLPis a unified lipophilicity and hydrophilicity potential. The interactions ofdipole and multipole moments, hydrogen bonds, and charged atoms in amolecule are included in the hydrophilic interactions in this model. TheHMLP is used to study hydrogen bonds and water–octanol partitioncoefficients in several examples. The calculated results show that the HMLPgives qualitatively and quantitatively correct, as well as chemicallyreasonable, results in cases where comparisons are available. Thesecomparisons indicate that the HMLP has advantages over the empiricallipophilicity potential in many aspects. The HMLP is a three-dimensional andeasily visualizable representation of molecular lipophilicity, suggested asa potential tool in computer-aided three-dimensional drug design.     

Determination of oil and water content in olive pomace using near infrared and Raman spectrometry. A comparative study

Near infrared (NIR) reflectance and Raman spectrometry were compared for determination of the oil and water content of olive pomace, a by-product in olive oil production. To enable comparison of the spectral techniques the same sample sets were used for calibration (1.74–3.93% oil, 48.3–67.0% water) and for validation (1.77–3.74% oil, 50.0–64.5% water). Several partial least squares (PLS) regression models were optimized by cross-validation with cancellation groups, including different spectral pretreatments for each technique. Best models were achieved with first-derivative spectra for both oil and water content. Prediction results for an independent validation set were similar for both techniques. The values of root mean square error of prediction (RMSEP) were 0.19 and 0.20–0.21 for oil content and 2.0 and 1.8 for water content, using Raman and NIR, respectively. The possibility of improving these results by combining the information of both techniques was also tested. The best models constructed using the appended spectra resulted in slightly better performance for oil content (RMSEP 0.17) but no improvement for water content.     

NIR imaging spectroscopy for quantification of constituents in polymers thin films loaded with paracetamol

Thin films loaded with the drug paracetamol were produced from polymer blends formed by hydroxypropylmethylcellulose (HPMC), polyvinylpyrrolidone (PVP) and polyethyleneglycol (PEG), at various mass ratios of polymers and drug defined by a d-optimal experimental design. NIR hyperspectral images were obtained from each thin film formulation and the pixel-to-pixel quantification of the constituents were carried out by partial least square (PLS) and multivariate curve resolution–alternating least square (MCR-ALS) with three different calibration/validation strategies. These strategies differ in the way to construct the calibration and validation matrices and they had to be carried out to suppress the bias on the quantification of the constituents in the polymer blend. The errors of prediction in the models from MCR-ALS were influenced by the calibration/validation strategy employed, but they were similar to the ones from PLS model. Concentration distribution maps were built after pixel-to-pixel predictions and their characteristics were analyzed.     

Successive projections algorithm improving the multivariate simultaneous direct spectrophotometric determination of five phenolic compounds in sea water

This paper proposes an analytical method to determine directly and simultaneously five phenolic compounds (4-nitrophenol, 2-nitrophenol, phenol, 2,4,6-trichlorophenol and 4-chlorophenol) in sea water (Ria de Bahía Blanca, Argentine). The advantages of this method is that only requires spectrophotometric measurements (separation steps and derivatization reagents are avoided) and chemometric modelling (PLS and MLR–SPA).The statistical comparison between PLS — a well established multivariate method — and MLR–SPA — a recently presented chemometric modelling — demonstrated better analytical performance for the later one. This fact is indicative of the potentiality of MLR–SPA for solving complex analytical problems.     

Predicting toxicity of benzene derivatives by molecular hologram derived quantitative structure–activity relationships (QSARS)

Holographic quantitative structure–activity relationship (HQSAR) is an emerging QSAR technique with the combined application of molecular hologram, which encodes the frequency of occurrence of various molecular fragment types, and the subsequent partial least squares (PLS) regression analysis. Based on molecular hologram, alignment-free QSAR models could be rapidly and easily developed with highly statistical significance and predictive ability. In this paper, the toxicity data for a series of 83 benzene derivatives to the autotrophic Chlorella vulgaris (IGC⁵⁰, negative logarithmic form of 6-h 50% population growth inhibition concentration in mmol/l) were subjected to HQSAR analysis and this resulted in a model with a high predictive ability. The robustness and predictive ability of the model were validated by “leave-one-out” (LOO) cross-validation procedure and an external testing set. The influence of fragment distinction parameters and fragment size on the quality of the HQSAR model have been also discussed.     

Research on anti-HIV-1 agents. Investigation on the CD4-Suradista binding mode through docking experiments

Sulfonated distamycin (Suradista) derivatives exhibit anti-HIV-1 activity by inhibiting the binding of the viral envelope glycoprotein gp120 to its receptor (CD4). With the aim to propose a possible binding mode between Suradistas and the CD4 macromolecule, molecular docking experiments, followed by energy minimization of the complexes thus obtained, were performed. Computational results show that ligand binding at the CD4 surface involves two or three positively charged regions of the macromolecule, in agreement with the results of X-ray crystallographic analysis of a ternary complex (CD4/gp120/neutralizing antibody) recently reported in the literature. Our findings account well for the structure–activity relationship found for Suradista compounds.     

Development of a unique 3D interaction model of endogenous and synthetic peripheral benzodiazepine receptor ligands

Different classes of Peripheral-type Benzodiazepine Receptor (PBR) ligands were examined and common structural elements were detected and used to develop a rational binding model based on energetically allowed ligand conformations. Two lipophilic regions and one electrostatic interaction site are essential features for high affinity ligand binding, while a further lipophilic region plays an important modulator role. A comparative molecular field analysis, performed over 130 PBR ligands by means of the GRID/GOLPE methodology, led to a PLS model with both high fitting and predictive values (r² = 0.898, Q² = 0.761). The outcome from the 3D QSAR model and the GRID interaction fields computed on the putative endogenous PBR ligands DBI (Diazepam Binding Inhibitor) and TTN (Tetracontatetraneuropeptide) was used to identify the amino acids most probably involved in PBR binding. Three amino acids, bearing lipophilic side chains, were detected in DBI (Phe49, Leu47 and Met46) and in TTN (Phe33, Leu31 and Met30) as likely residues underlying receptor binding. Moreover, a qualitative comparison of the molecular electrostatic potentials of DBI, TTN and selected synthetic ligands indicated also similar electronic properties. Convergent results from the modeling studies of synthetic and endogenous ligands suggest a common binding mode to PBRs. This may help the rational design of new high affinity PBR ligands.