Alla P. Toropova

CORAL: Quantitative structure–activity relationship models for estimating toxicity of organic compounds in rats

For six random splits, one‐variable models of rat toxicity (minus decimal logarithm of the 50% lethal dose [pLD50], oral exposure) have been calculated with CORAL software (http://www.insilico.eu/coral/). The total number of considered compounds is 689. New additional global attributes of the simplified molecular input line entry system (SMILES) have been examined for improvement of the optimal SMILES‐based descriptors. These global SMILES attributes are representing the presence of some chemical elements and different kinds of chemical bonds (double, triple, and stereochemical). The “classic” scheme of building up quantitative structure–property/activity relationships and the balance of correlations (BC) with the ideal slopes were compared. For all six random splits, best prediction takes place if the aforementioned BC along with the global SMILES attributes are included in the modeling process. The average statistical characteristics for the external test set are the following: n = 119 ± 6.4, R2 = 0.7371 ± 0.013, and root mean square error = 0.360 ± 0.037.

Novel application of the CORAL software to model cytotoxicity of metal oxide nanoparticles to bacteria Escherichia coli.

Convenient to apply and available on the Internet software CORAL (http://www.insilico.eu/CORAL) has been used to build up quantitative structure-activity relationships (QSAR) for prediction of cytotoxicity of metal oxide nanoparticles to bacteria Escherichia coli (minus logarithm of concentration for 50% effect pEC50). In this study six random splits of the data into the training and test set were examined. It has been shown that the CORAL provides a reliable tool that could be used to build up a QSAR of the pEC50.

Prediction of heteroaromatic amine mutagenicity by means of correlation weighting of atomic orbital graphs of local invariants

Abstract Graphs of atomic orbitals (GAOs) have been used to represent molecular structures. Rules by which the labeled hydrogen-filled graphs (LHFGs) were converted into the GAOs are described. The GAO is an attempt at taking into account the structures of atoms (i.e. atomic orbitals, such as 1s 1 , 2p 2 , 3d 10 ) for QSPR/QSAR analyses. As a method of mutagenicity modeling, optimization of correlation weights of local invariants (OCWLI) of the LHFGs and the GAOs has been used. Statistical characteristics of such models based on the OCWLI of GAO are better than those based on the OCWLI of the LHFGs.

QSAR modeling of toxicity on optimization of correlation weights of Morgan extended connectivity

Abstract Optimal descriptors calculated in the presence of correlation weights in molecular graph of different kinds of vertices (i.e. chemical elements, e.g. H, C, O, etc.) and different Morgan extended connectivity of third-order values are reasonably better models of the acute aquatic toxicity (−log[LC 50 ]) of benzene derivatives. Statistical characteristics of the model are the following: on training set n =44, r 2 =0.8982, s =0.251, F =371; on test set n =25, r 2 =0.9181, s =0.234, F =258.

Quasi-QSAR for mutagenic potential of multi-walled carbon-nanotubes

Available on the Internet, the CORAL software (http://www.insilico.eu/coral) has been used to build up quasi-quantitative structure-activity relationships (quasi-QSAR) for prediction of mutagenic potential of multi-walled carbon-nanotubes (MWCNTs). In contrast with the previous models built up by CORAL which were based on representation of the molecular structure by simplified molecular input-line entry system (SMILES) the quasi-QSARs based on the representation of conditions (not on the molecular structure) such as concentration, presence (absence) S9 mix, the using (or without the using) of preincubation were encoded by so-called quasi-SMILES. The statistical characteristics of these models (quasi-QSARs) for three random splits into the visible training set and test set and invisible validation set are the following: (i) split 1: n=13, r(2)=0.8037, q(2)=0.7260, s=0.033, F=45 (training set); n=5, r(2)=0.9102, s=0.071 (test set); n=6, r(2)=0.7627, s=0.044 (validation set); (ii) split 2: n=13, r(2)=0.6446, q(2)=0.4733, s=0.045, F=20 (training set); n=5, r(2)=0.6785, s=0.054 (test set); n=6, r(2)=0.9593, s=0.032 (validation set); and (iii) n=14, r(2)=0.8087, q(2)=0.6975, s=0.026, F=51 (training set); n=5, r(2)=0.9453, s=0.074 (test set); n=5, r(2)=0.8951, s=0.052 (validation set).

SMILES-based optimal descriptors: QSAR analysis of fullerene-based HIV-1 PR inhibitors by means of balance of correlations

Quantitative structure‐activity relationships (QSAR) for prediction of binding affinities (pEC50, i.e., minus decimal logarithm of the 50% effective concentration) of 20 fullerene derivatives inhibitors of the HIV‐1 PR (human immunodeficiency virus type 1 protease) have been developed by application of the optimal descriptors approach calculated with SMILES (simplified molecular input line entry system). The applied models were constructed by the balance of correlations. Three various splits of the experimental data into subtraining set, calibration set, and test set were examined. Comparison of classic scheme (training‐test system) and the balance of correlations (subtraining‐calibration‐test system) show that the balance of correlations gives more robust predictions than the classic scheme for the pEC50 of the fullerene derivatives.

CORAL: Building up the model for bioconcentration factor and defining it's applicability domain

CORAL (CORrelation And Logic) software can be used to build up the quantitative structure--property/activity relationships (QSPR/QSAR) with optimal descriptors calculated with the simplified molecular input line entry system (SMILES). We used CORAL to evaluate the applicability domain of the QSAR models, taking a model of bioconcentration factor (logBCF) as example. This models based on a large training set of more than 1000 chemicals. To improve the model is predictivity and reliability on new compounds, we introduced a new function, which uses the Delta(obs) = logBCF(expr)--logBCF(calc) of the predictions on the chemicals in the training set. With this approach, outliers are eliminated from the phase of training. This proved useful and increased the models predictivity.

Optimal descriptor as a translator of eclectic data into endpoint prediction: Mutagenicity of fullerene as a mathematical function of conditions

The experimental data on the bacterial reverse mutation test on C60 nanoparticles (TA100) is examined as an endpoint. By means of the optimal descriptors calculated with the Monte Carlo method a mathematical model of the endpoint has been built up. The model is the mathematical function of (i) dose (g/plate); (ii) metabolic activation (i.e. with S9 mix or without S9 mix); and (iii) illumination (i.e. dark or irradiation). The statistical quality of the model is the following: n=10, r(2)=0.7549, q(2)=0.5709, s=7.67, F=25 (Training set); n=5, r(2)=0.8987, s=18.4 (Calibration set); and n=5, r(2)=0.6968, s=10.9 (Validation set).

Modeling of acyclic carbonyl compounds normal boiling points by correlation weighting of nearest neighboring codes

Abstract The nearest neighboring code (NNC) is a local graph invariant. The NNC of a given vertex of the labeled hydrogen filled graph (LHFG) is a function of atom composition of the vertex neighbors. By optimization the correlation weights of different atoms and different values of the NNCs, one-variable models of the normal boiling points of carbonyl compounds have been obtained. Statistical characteristics of the best of such model are as follows: training set: n =100, r 2 =0.972, s =6.12°C, F =3464; test set: n =100, r 2 =0.975, s =6.00°C, F =3905.

3D weighting of molecular descriptors for QSPR/QSAR by the method of ideal symmetry (MIS). 1. Application to boiling points of alkanes

Abstract The method of ideal symmetry (MIS), developed recently, presents molecules as systems of mutually repulsing atoms connected by covalent bonds of constant length. In this paper we have used MIS optimized geometry to define a vertex 3D weight as a metric analogue of the vertex distance sum in molecular graphs. These 3D weights were used as a substitute for the vertex degrees in several well known topological (2D) indices, thus producing a series of 3D-weighted molecular descriptors. The novel indices were tested in calculating the boiling points of a series of 73 C3C9 alkanes and showed generally a better performance than the original 2D indices. The best 1-, 2-, and 3-variable linear regression models incorporated 3D zero-order molecular connectivity with correlation coefficients of 0.9892, 0.9961, and 0.9986, and standard deviations of 5.97, 3.64, and 2.17 °C, respectively. The approach was further validated by correlations with four other properties of alkanes (heats of formation, heats of vaporization, heats of atomization, and molar volume). The potential of the proposed 3D weighting of topological indices for QSPR/QSAR studies was thus demonstrated.