Infrared Spectra Simulation of Substituted Benzene Derivatives on the Basis of a 3D Structure Representation

A Concise Review on Recent Developments of Machine Learning for the Prediction of Vibrational Spectra

Machine learning has become more and more popular in computational chemistry, as well as in the important field of spectroscopy. In this concise review, we walk the reader through a short summary of machine learning algorithms and a comprehensive discussion on the connection between machine learning methods and vibrational spectroscopy, particularly for the case of infrared and Raman spectroscopy. We also briefly discuss state-of-the-art molecular representations which serve as meaningful inputs for machine learning to predict vibrational spectra. In addition, this review provides an overview of the transferability and best practices of machine learning in the prediction of vibrational spectra as well as possible future research directions.

Computer-aided Discovery of Peptides that Specifically Attack Bacterial Biofilms

Biofilms represent a multicellular growth state of bacteria that are intrinsically resistant to conventional antibiotics. It was recently shown that a synthetic immunomodulatory cationic peptide, 1018 (VRLIVAVRIWRR-NH₂), exhibits broad-spectrum antibiofilm activity but the sequence determinants of antibiofilm peptides have not been systematically studied. In the present work, a peptide library consisting of 96 single amino acid substituted variants of 1018 was SPOT-synthesized on cellulose arrays and evaluated against methicillin resistant Staphylococcus aureus (MRSA) biofilms. This dataset was used to establish quantitative structure-activity relationship (QSAR) models relating the antibiofilm activity of these peptides to hundreds of molecular descriptors derived from their sequences. The developed 3D QSAR models then predicted the probability that a peptide would possess antibiofilm activity from a library of 100,000 virtual peptide sequences in silico. A subset of these variants were SPOT-synthesized and their activity assessed, revealing that the QSAR models resulted in ~85% prediction accuracy. Notably, peptide 3002 (ILVRWIRWRIQW-NH₂) was identified that exhibited an 8-fold increased antibiofilm potency in vitro compared to 1018 and proved effective in vivo, significantly reducing abscess size in a chronic MRSA mouse infection model. This study demonstrates that QSAR modeling can successfully be used to identify antibiofilm specific peptides with therapeutic potential.

A quantitative structure–property study of reorganization energy for known p-type organic semiconductors

Intramolecular reorganization energy (RE), which quantifies the electron-phonon coupling strength, is an important charge transport parameter for the theoretical characterization of molecular organic semiconductors (OSCs). On a small scale, the accurate calculation of the RE is trivial; however, for large-scale screening, faster approaches are desirable. We investigate the structure–property relations and present a quantitative structure–property relationship study to facilitate the computation of RE from molecular structure. To this end, we generated a compound set of 171, which was derived from known p-type OSCs built from moieties such as acenes, thiophenes, and pentalenes. We show that simple structural descriptors such as the number of atoms, rings or rotatable bonds only weakly correlate with the RE. On the other hand, we show that regression models based on a more comprehensive representation of the molecules such as SMILES-based molecular signatures and geometry-based molecular transforms can predict the RE with a coefficient of determination of 0.7 and a mean absolute error of 40 meV in the library, in which the RE ranges from 76 to 480 meV. Our analysis indicates that a more extensive compound set for training is necessary for more predictive models.

An investigation of the structure–property relationship between reorganization energy and molecular structure.

Chemoinformatics: Achievements and Challenges, a Personal View

Chemoinformatics provides computer methods for learning from chemical data and for modeling tasks a chemist is facing. The field has evolved in the past 50 years and has substantially shaped how chemical research is performed by providing access to chemical information on a scale unattainable by traditional methods. Many physical, chemical and biological data have been predicted from structural data. For the early phases of drug design, methods have been developed that are used in all major pharmaceutical companies. However, all domains of chemistry can benefit from chemoinformatics methods; many areas that are not yet well developed, but could substantially gain from the use of chemoinformatics methods. The quality of data is of crucial importance for successful results. Computer-assisted structure elucidation and computer-assisted synthesis design have been attempted in the early years of chemoinformatics. Because of the importance of these fields to the chemist, new approaches should be made with better hardware and software techniques. Society's concern about the impact of chemicals on human health and the environment could be met by the development of methods for toxicity prediction and risk assessment. In conjunction with bioinformatics, our understanding of the events in living organisms could be deepened and, thus, novel strategies for curing diseases developed. With so many challenging tasks awaiting solutions, the future is bright for chemoinformatics.

3D-MoRSE descriptors explained

3D-MoRSE is a very flexible 3D structure encoding framework for chemoinformatics and QSAR purposes due to the range of scattering parameter values and variety of weighting schemes used. While arising in many QSAR studies, up to this time they were considered as hardly interpreted and were treated like a "black box". This study is intended to lift the veil of mystery, providing a comprehensible way to the interpretation of 3D-MoRSE descriptors in QSAR/QSPR studies. The values of these descriptors are calculated with rather simple equation, but may vary when using differing starting geometries as optimization input. This variation increases with scattering parameter and also is higher for electronegativity weighted and unweighted descriptors. Though each 3D-MoRSE descriptor incorporates the information about the whole molecule structure, its final value is derived mostly from short-distance (up to 3Å) atomic pairs. And, if a QSAR study covers structurally similar set of compounds, then the role of 3D-MoRSE descriptor in a model can be interpreted using just several pairs of neighbor atoms. The guide to interpretation process is discussed and illustrated with a case study. Realizing the mathematical concept behind 3D-descriptors and knowing their properties it is easy not only to interpret, but also to predict the importance of 3D-MoRSE descriptors in a QSAR study. The process of prediction is described on the practical example and its accuracy is confirmed with further QSAR modeling.

Spherical harmonics coefficients for ligand-based virtual screening of cyclooxygenase inhibitors

BACKGROUND

Molecular descriptors are essential for many applications in computational chemistry, such as ligand-based similarity searching. Spherical harmonics have previously been suggested as comprehensive descriptors of molecular structure and properties. We investigate a spherical harmonics descriptor for shape-based virtual screening.

METHODOLOGY/PRINCIPAL FINDINGS

We introduce and validate a partially rotation-invariant three-dimensional molecular shape descriptor based on the norm of spherical harmonics expansion coefficients. Using this molecular representation, we parameterize molecular surfaces, i.e., isosurfaces of spatial molecular property distributions. We validate the shape descriptor in a comprehensive retrospective virtual screening experiment. In a prospective study, we virtually screen a large compound library for cyclooxygenase inhibitors, using a self-organizing map as a pre-filter and the shape descriptor for candidate prioritization.

CONCLUSIONS/SIGNIFICANCE

12 compounds were tested in vitro for direct enzyme inhibition and in a whole blood assay. Active compounds containing a triazole scaffold were identified as direct cyclooxygenase-1 inhibitors. This outcome corroborates the usefulness of spherical harmonics for representation of molecular shape in virtual screening of large compound collections. The combination of pharmacophore and shape-based filtering of screening candidates proved to be a straightforward approach to finding novel bioactive chemotypes with minimal experimental effort.

Modeling the excitation wavelengths (lambda(ex)) of boronic acids

The quantitative structure-property relationship (QSPR) method was used to model the fluorescence excitation wavelengths (lambda(ex)) of 42 boronic acid-based fluorescent biosensors (30 in the training set and 12 in the test set). In this QSPR study, unsupervised forward selection (UFS), stepwise multiple linear regression (SMLR), partial least squares regression (PLS) and associative neural networks (ASNN) were employed to simulate linear and nonlinear models. All models were validated by a test set and Tropsha's validation model. The resulting ASNN nonlinear model demonstrates significant improvement on the predictive ability of the neural network compared to the SMLR and PLS linear models. The descriptors used in the models are discussed in detail. These QSPR models are useful tools for the prediction of fluorescence excitation wavelengths of arylboronic acids.

N-[2-(5,5-Dimethyl-1,3-dioxane-2-yl)ethyl]amino acids: Their synthesis, anti-inflammatory evaluation and QSAR analysis

Developing novel anti-inflammatory drugs is increasingly important in modern pharmaceutical industry. In this work, the reactions of both amino acids and their methylesters with 3-(5,5-dimethyl-1,3-dioxane-2-yl)propanal (2) were performed to either directly provide the goal products N-[2-(5,5-dimethyl-1,3-dioxane-2-yl)ethyl]amino acids (4a-s) in 9-65% yields or provide the intermediates N-[2-(5,5-dimethyl-1,3-dioxane-2-yl)ethyl]amino acid methylesters (3a-s) in 78-87% yields. The saponification of 3a-s provided 4a-s in 80-89% yields. Using a xylene-induced ear edema model, the anti-inflammatory activities of these newly synthesized anti-inflammatory agents were evaluated. The results indicated that comparing to the vehicle control 17 out of 4a-s significantly inhibited the development of inflammation in mice (p<0.01). In particular, eight out of 4a-s exhibited an even higher anti-inflammatory activity than the standard reference drug aspirin (p<0.05-0.01). A QSAR analysis was performed by use of the molecular descriptors generated from e-dragon software. The predictive accuracy of the established QSAR model implies that it may be promising for screening the new derivatives of 2-position amino acid substituted 1,3-dioxanes as potential anti-inflammatory agents.

Exploration of physicochemical properties and molecular modelling studies of 2-sulfonyl-phenyl-3-phenyl-indole analogs as cyclooxygenase-2 inhibitors

In the present work, modelling study has been performed to explore the physicochemical requirements of 2-sulfonyl-phenyl-3-phenyl-indole analogs as COX-2 enzyme inhibitors. The multivariant regression expressions were developed using sequential multiple linear regression (SEQ-MLR) technique, considering adjustable correlation coefficient (r(adj)(2)). The statistical quality of SEQ-MLR equations was evaluated considering parameters like correlation coefficient (r), standard error of estimation (SEE), and variance ratio (F) at explicit degree of freedom (df). Orthogonality of the descriptors in SEQ-MLR was established through variance inflation factor (VIF). Developed equations have been internally validated using leave-one-out technique and further validated with test set, considering predictive squared correlation coefficient (r(pred)(2)). The orientation of the most potent and selective COX-2 inhibitor of training set, 2-(4-phenyl sulfonamide)-3-phenyl-5-methylindole, in the COX-2 active site was explored by docking. The phenyl sulfonamide moiety positioned in secondary pocket of enzyme which consists of amino acid residues Phe(518), Gln(192), Arg(513), Leu(352), Ser(353) and Val(523) is responsible for the selectivity. The unsubstituted phenyl ring positions in a hydrophobic cavity are lined by Tyr(385), Trp(387), Tyr(348), Leu(384) and Met(522). Interestingly, the indole C-5 CH(3)-substituent is located in a hydrophobic region formed by Ile(345), Val(349), Ala(527), Leu(531) and Leu(534). The hydrophobic interactions of methyl group might be crucial for the potency of 2-sulfonyl-phenyl-3-phenyl-indole analogs. Study has revealed that atomic van der Waals volume and atomic masses explain COX-2 inhibitory activity of 2-sulfonyl-phenyl-3-phenyl-indole analogs significantly.

Molecular modeling of histamine H3 receptor and QSAR studies on arylbenzofuran derived H3 antagonists

Histamine H3 receptors are presynaptic autoreceptors found in both central and peripheral nervous systems of many species. The central effects of these receptors suggest a potential therapeutic role for their antagonists in treatment of several neurological disorders such as epilepsy, schizophrenia, Alzheimer's and Parkinson's diseases. The purpose of this study was to identify the structural requirements for H3 antagonistic activity via quantitative structure-activity relationship (QSAR) studies and receptor modeling/docking techniques. A combination of partial least squares (PLS) and genetic algorithm (GA) was used in the QSAR approach to select the structural descriptors relevant to the receptor binding affinity of a series of 58 H3 antagonists. The descriptors were selected out of a pool of >1000 descriptors calculated by DRAGON, Hyperchem and ACD labs suite of programs. The resulting QSAR models for rat and human H3 binding affinities were validated using different strategies. QSAR models generated in the current work suggested the role of charge transfer interactions in the ligand-receptor interaction verified using the molecular modeling of the receptor and docking two antagonists to the binding site. The 3D model of human H3 receptor was built based on bovine rhodopsin structure and evaluated by molecular dynamics (MD) simulation in a mixed water-vacuum-water environment. The results were indicative of the stability of the model relating the observed structural changes during the MD simulation to the suggested ligand-receptor interactions. The results of this investigation are expected to be useful in the process of design and development of new potent H3 receptor antagonists.

Rationalization of physicochemical characters of oxazolyl thiosemicarbazone analogs towards multi-drug resistant tuberculosis: a QSAR approach

The emergence of multi-drug resistant (MDR) strains of Mycobacterium tuberculosis and the continuing pandemic of tuberculosis emphasizes the urgent need for the development of new and potent anti-tubercular agents. In an effort to develop new and more effective agents to treat tuberculosis emphasis was focused on quantification of structure-activity relationship of oxazolyl thiosemicarbazone derivatives. The de novo analysis gave insight to some important structural features i.e. nitro group on phenyl ring at R(1) position is optimal for the activity and might be responsible for electronic interaction, while phenyl ring at R position interact with the hydrophobic pocket more effectively as compared to unsubstituted or methyl substituted analogs. Hansch approach offered the understanding and parameterization of interactions of the inhibitor with receptor. Similarly QSAR analysis gave some important physicochemical properties, i.e. empirical aromatic index (ARR) and 3D-MoRSE code value of scattering angle at 8A(-1). These two physicochemical properties shall be helpful in the development of more potent analogs.

Simultaneous determination of thiocyanate and salicylate by a combined UV-spectrophotometric detection principal component artificial neural network

A modified principle component artificial neural network (PC-ANN) model is developed for simultaneous determination of thiocyanate and salycilate concentration after passing through the bulk of a liquid membrane by tri-phenyl benzyl phosphonium chloride. All calibration, and test samples data were obtained using UV-Vis spectrophotometer. In this way, a modified PC-ANN consisting of three layers of nodes was trained by combination of Bayesian-Levenberg-Marquardt as training rule. Sigmoid and liner transfer functions were used in the hidden and output layers respectively to facilitate nonlinear calibration. The model could accurately estimate the concentration of components with acceptable precision and accuracy, for mixtures. The PC-ANN model exhibits a good ability for the simultaneous determination of the thiocyanate and salycilate in concentration range 0.5 x 10(-4) mol.l(-1) up to 5.0 x 10(-4) mol.l(-1) with Root Mean square error (2.22% and 2.20%, for thiocyanate and salycilate, respectively) and high correlation coefficients (R2= 0.998 or greater). Results obtained with modified trained PC-ANN were compared with stepwise linear regression (SMLR) model. Validation of the two models shows a better ability in estimation of the modified PC-ANN as compared with the SMLR model (MSRE given are 3.12%, 6.31%.).

Rationalization of physicochemical characters of 1,5-diarylpyrazole analogs as dual (COX-2/LOX-5) inhibitors: A QSAR approach

Arachidonic acid metabolizing enzymes cyclooxygenases and lipoxygenases lead to the formation of various eicosanoids involved in variety of human diseases, like inflammation, fever, pain, rheumatic and osteoarthritis, etc. Non-steroidal anti-inflammatory drugs are useful tools in the treatment of prostaglandin mediated complications. The development of dual inhibitors may prevent a drift of arachidonic acid metabolism towards the other pathway, leading to potential side effects. Hence emphasis was focused on quantification of structure-activity relationship, with a view to delineating the influence of key COX-2/LOX-5 activity, to explore structural insights to aid the designing of safer dual inhibitors. The quantification of the structural features of 1,5-diarylpyrazole analogs with various biological activities gave some important structural insights, i.e. Hy (hydrophilic factor) and Mor17v (3D molecular representation of structure based on electron diffraction code). These two physicochemical properties may be helpful in development of more selective dual COX-2/LOX-5 inhibitors.

Validation of a QSAR model for acute toxicity

In the present study, a quantitative structure--activity relationship (QSAR) model has been developed for predicting acute toxicity to the fathead minnow (Pimephales promelas), the aim being to demonstrate how statistical validation and domain definition are both required to establish model validity and to provide reliable predictions. A dataset of 408 heterogeneous chemicals was modelled by a diverse set of theoretical molecular descriptors by using multivariate linear regression (MLR) and Genetic Algorithm-Variable Subset Selection (GA-VSS). This QSAR model was developed to generate reliable predictions of toxicity for organic chemicals not yet tested, so particular emphasis was given to statistical validity and applicability domain. External validation was performed by using OECD Screening Information Data Set (SIDS) data for 177 High Production Volume (HPV) chemicals, and a good predictivity was obtained (=72.1). The model was evaluated according to the OECD principles for QSAR validation, and compliance with all five principles was established. The model could therefore be useful for the regulatory assessment of chemicals. For example, it could be used to fill data gaps within its chemical domain and contribute to the prioritization of chemicals for aquatic toxicity testing.

Total ranking models by the genetic algorithm variable subset selection (GA?VSS) approach for environmental priority settings

Total order ranking (TOR) strategies, which are mathematically based on elementary methods of discrete mathematics, seem to be attractive and simple tools for performing data analysis. Moreover order-ranking strategies seem to be a very useful tool not only to perform data exploration but also to develop order ranking models, a possible alternative to conventional quantitative structure-activity relationship (QSAR) methods. In fact, when data material is characterised by uncertainties, order methods can be used as alternative to statistical methods such as multilinear regression (MLR), because they do not require specific functional relationships between the independent and dependent variables (responses). A ranking model is a relationship between a set of dependent attributes, experimentally investigated, and a set of independent attributes, i.e. model attributes, which are calculated attributes. As in regression and classification models, the variable selection model is one of the main steps in finding predictive models. In this work the genetic algorithm-variable subset selection (GA-VSS) approach is proposed as the variable selection method for searching for the best ranking models within a wide set of variables. The models based on the selected subsets of variables are compared with the experimental ranking and evaluated by the Spearman's rank index. A case study application is presented on a TOR model developed for polychlorinated biphenyl (PCB) compounds, which have been analysed according to some of their physicochemical properties which play an important role in their environmental impact.

Prediction of aromatic amines mutagenicity from theoretical molecular descriptors

In the present research the mutagenicity data (Ames tests TA98 and TA100) for various aromatic and heteroaromatic amines, a data set extensively studied by other quantitative structure-activity relationship (QSAR)-authors, have been modeled by a wide set of theoretical molecular descriptors using linear multivariate regression (MLR) and genetic algorithm-variable subset selection (GA-VSS). The models have been calculated on a subset of compounds selected by a D-optimal experimental design. Moreover, they have been validated by both internal and external validation procedures showing satisfactory predictive performance. The models proposed here can be useful in predicting data and setting a testing priority for those compounds for which experimental data are not available or are not yet synthesized.

Quantitative structure-property relationship studies of migration index in microemulsion electrokinetic chromatography using artificial neural network

Artificial neural networks (ANNs) were successfully developed for the modeling and prediction of migration indices of the 53 benzene derivatives and heterocyclic compounds in microemulsion electrokinetic chromatography. The selected descriptors that appear in multiple linear regression models are: 3D-MoRSE signal 25 unweighted, 3D-MoRSE signal 19 weighted by atomic Sanderson electronegativity, R maximal autocorrelation index lag 1 weighted by atomic mass (R1M+), R maximal autocorrelation index lag 2 weighted by polarizability (R2P+) and average atomic composition index. These descriptors were used as inputs for generated 5-4-1 networks. After training and optimization of the ANN parameters it was used to prediction of migration index of the test set compounds. The results obtained using ANNs were compared with the experimental values as well as with those obtained using regression models and showed the superiority of ANNs over regression models.

Fourier transform infrared attenuated total reflection and transmission spectra studied by dispersion analysis

Fourier transform infrared transmission (FT-IR) and attenuated total reflection (ATR) spectra of water-ethanol mixtures are recorded and reconstructed thanks to a causal dispersion analysis technique. As expected, the Beer's law technique is an empirical approximate method that cannot account for complex spectral features. On the other hand, a rigorous analysis performed by using the theoretical optical paths for both experimental techniques and Gaussian dispersion analysis (GDA) allows the dielectric functions of the pure liquids to be calculated. Simulations of the whole mid-infrared spectra in the range 500-4000 cm(-1) match the experimental data very well, whatever the water-ethanol mixtures. This method is a powerful tool to quantify such model mixtures and more generally could be the first step toward software for assistance to the FT-IR spectrum analysis.

Structure/response correlations and similarity/diversity analysis by GETAWAY descriptors. 1. Theory of the novel 3D molecular descriptors

Novel molecular descriptors based on a leverage matrix similar to that defined in statistics and usually used for regression diagnostics are presented. This leverage matrix, called Molecular Influence Matrix (MIM), is here proposed as a new molecular representation easily calculated from the spatial coordinates of the molecule atoms in a chosen conformation. The proposed molecular descriptors are called GETAWAY (GEometry, Topology, and Atom-Weights AssemblY) as they try to match 3D-molecular geometry provided by the molecular influence matrix and atom relatedness by molecular topology, with chemical information by using different atomic weightings (atomic mass, polarizability, van der Waals volume, and electronegativity, together with unit weights). A first set of molecular descriptors, called H-GETAWAY, is derived by using only the information provided by the molecular influence matrix, while a second set, called R-GETAWAY, combines this information with geometric interatomic distances in the molecule. The prediction ability in structure-property correlations of the new descriptors was tested by analyzing regressions of these descriptors for selected properties of octanes.

Decision support systems for chemical structure representation, reaction modeling, and spectra simulation

The choice of an appropriate structure coding scheme is the secret to success in QSAR studies. Depending on the problem at hand, 2D or 3D descriptors have to be chosen; the consideration of electronic effects might be crucial, conformational flexibility has to be of special concern. Artificial neural networks, both with unsupervised and with supervised learning schemes, are powerful tools for establishing relationships between structure and physical, chemical, or biological properties. The EROS system for the simulation of chemical reactions is briefly presented and its application to the degradation of s-triazine herbicides is shown. It is further shown how the simulation of chemical reactions can be combined with the simulation of infrared spectra for the efficient identification of the structure of degradation products.

SIRS-SS: a system for simulating IR/Raman spectra. 2. Procedures and performance

This paper is devoted to the description of procedures used in our IR/RAMAN spectrum simulation system, based on substructure/subspectrum correlations established between linked databases. The search is performed in the following order: small molecules/specific fragments/atom centered FRELs (FREL: FRagment centered on an Environment which is Limited)/bond focused FRELs. Comparative study with several reported methods has been carried out to show good performance of this software both for IR and RAMAN spectrum simulation.

SIRS-SS: a system for simulating IR/Raman spectra. 1. Substructure/subspectrum correlation

An IR/RAMAN spectra simulation system is reported. The development of this software was based on the substructure/subspectrum relationships established for four different structural classes: small molecules, special fragments, atom-centered FRELs, and bond-centered FRELs (FREL: Fragment centered on an Environment which is Limited). Four corresponding knowledge-bases (now, at a pilot stage) are constructed from usual correlation charts or data analyses of large populations of compounds using data mining techniques.

A combined application of reaction prediction and infrared spectra simulation for the identification of degradation products of s-triazine herbicides

Substance identification in analytical chemistry is usually performed by comparing an experimental spectrum with a reference spectrum. Especially in environmental chemistry, reference spectra from databases are only available for a limited number of compounds. The combination of the reaction prediction system EROS and of infrared spectra simulation is a powerful tool for computer-assisted substance identification. First, possible degradation products of a chemical are predicted and then the infrared spectra of all these compounds are simulated. Comparison of the simulated infrared spectra with experimental spectra allows one to identify the structure of compounds. The method is demonstrated with the example of s-triazine herbicides.

Automatic generation of knowledge base from infrared spectral database for substructure recognition

This paper presents a new methodology of chemical substructure recognition by interpretation of an infrared spectrum. The approach in spectrum interpretation is based on the determination of functional groups, which may be present or absent in compounds whose structure is unknown. The process of searching for spectrum-substructure correlation is realized by application of a statistical algorithm. In this method, correlations are generalized and condensed into a set of interpretation rules which are applied to the interpretation of an unknown compound's spectrum in order to predict whether the respective substructures are present or absent in the unknown molecule.