RESUMO
Cluster tendency assessment is an important stage in cluster analysis. In this sense, a group of promising techniques named visual assessment of tendency (VAT) has emerged in the literature. The presence of clusters can be detected easily through the direct observation of a dark blocks structure along the main diagonal of the intensity image. Alternatively, if the Dunn's index for a single linkage partition is greater than 1, then it is a good indication of the blocklike structure. In this report, the Dunn's index is applied as a novel measure of tendency on 8 pharmacological data sets, represented by machine-learning-selected molecular descriptors. In all cases, observed values are less than 1, thus indicating a weak tendency for data to form compact clusters. Other results suggest that there is an increasing relationship between the Dunn's index as a measure of cluster separability and the classification accuracy of various cluster algorithms tested on the same data sets.
Assuntos
Análise por Conglomerados , Interpretação Estatística de Dados , Bases de Dados Factuais/estatística & dados numéricos , Farmacologia/estatística & dados numéricos , Humanos , SoftwareRESUMO
Cluster algorithms play an important role in diversity related tasks of modern chemoinformatics, with the widest applications being in pharmaceutical industry drug discovery programs. The performance of these grouping strategies depends on various factors such as molecular representation, mathematical method, algorithmical technique, and statistical distribution of data. For this reason, introduction and comparison of new methods are necessary in order to find the model that best fits the problem at hand. Earlier comparative studies report on Ward's algorithm using fingerprints for molecular description as generally superior in this field. However, problems still remain, i.e., other types of numerical descriptions have been little exploited, current descriptors selection strategy is trial and error-driven, and no previous comparative studies considering a broader domain of the combinatorial methods in grouping chemoinformatic data sets have been conducted. In this work, a comparison between combinatorial methods is performed,with five of them being novel in cheminformatics. The experiments are carried out using eight data sets that are well established and validated in the medical chemistry literature. Each drug data set was represented by real molecular descriptors selected by machine learning techniques, which are consistent with the neighborhood principle. Statistical analysis of the results demonstrates that pharmacological activities of the eight data sets can be modeled with a few of families with 2D and 3D molecular descriptors, avoiding classification problems associated with the presence of nonrelevant features. Three out of five of the proposed cluster algorithms show superior performance over most classical algorithms and are similar (or slightly superior in the most optimistic sense) to Ward's algorithm. The usefulness of these algorithms is also assessed in a comparative experiment to potent QSAR and machine learning classifiers, where they perform similarly in some cases.
Assuntos
Modelos Estatísticos , Relação Quantitativa Estrutura-Atividade , Algoritmos , Inteligência Artificial , Análise por Conglomerados , Modelos Biológicos , Preparações Farmacêuticas/química , FarmacologiaRESUMO
The interaction of three different brassinosteroids with water was studied by the Multiple Minima Hypersurface (MMH) procedure to model molecular interactions explicitly. The resulting thermodynamic data give useful information on properties of molecular association with water. This application can serve as a tool for future investigations and modelling concerning interactions of brassinosteroids with receptor proteins in plants. DFT/B3LYP calculations were also made in order to correlate and test the performance of the current AM1 Hamiltonian calculations of these complexes, which are inherent to MMH routine. Diol functionalities located in ring A and lateral chain appears as the sites that show the highest affinity to water. The oxalactone group does not appear to be a key structural requirement in the association with water. Parallel calculations with a "polarizable continuum method" (PCM) agreed with the reported experimental order of biological activities, where Brassinolide exhibited the best solubility features.
Assuntos
Colestanóis/química , Modelos Teóricos , Reguladores de Crescimento de Plantas/química , Esteroides Heterocíclicos/química , Água/química , Brassinosteroides , Estrutura Molecular , Solubilidade , TermodinâmicaRESUMO
Predictive quantitative structure-activity relationship (QSAR) models of anabolic and androgenic activities for the testosterone and dihydrotestosterone steroid analogues were obtained by means of multiple linear regression using quantum and physicochemical molecular descriptors (MD) as well as a genetic algorithm for the selection of the best subset of variables. Quantitative models found for describing the anabolic (androgenic) activity are significant from a statistical point of view: R(2) of 0.84 (0.72 and 0.70). A leave-one-out cross-validation procedure revealed that the regression models had a fairly good predictability [q(2) of 0.80 (0.60 and 0.59)]. In addition, other QSAR models were developed to predict anabolic/androgenic (A/A) ratios and the best regression equation explains 68% of the variance for the experimental values of AA ratio and has a rather adequate q(2) of 0.51. External validation, by using test sets, was also used in each experiment in order to evaluate the predictive power of the obtained models. The result shows that these QSARs have quite good predictive abilities (R(2) of 0.90, 0.72 (0.55), and 0.53) for anabolic activity, androgenic activity, and A/A ratios, respectively. Last, a Williams plot was used in order to define the domain of applicability of the models as a squared area within +/-2 band for residuals and a leverage threshold of h=0.16. No apparent outliers were detected and the models can be used with high accuracy in this applicability domain. MDs included in our QSAR models allow the structural interpretation of the biological process, evidencing the main role of the shape of molecules, hydrophobicity, and electronic properties. Attempts were made to include lipophilicity (octanol-water partition coefficient (logP)) and electronic (hardness (eta)) values of the whole molecules in the multivariate relations. It was found from the study that the logP of molecules has positive contribution to the anabolic and androgenic activities and high values of eta produce unfavorable effects. The found MDs can also be efficiently used in similarity studies based on cluster analysis. Our model for the anabolic/androgenic ratio (expressed by weight of levator ani muscle, LA, and seminal vesicle, SV, in mice) predicts that the 2-aminomethylene-17alpha-methyl-17beta-hydroxy-5alpha-androstan-3-one (43) compound is the most potent anabolic steroid, and the 17alpha-methyl-2beta,17beta-dihydroxy-5alpha-androstane (31) compound is the least potent one of this series. The approach described in this report is an alternative for the discovery and optimization of leading anabolic compounds among steroids and analogues. It also gives an important role to electron exchange terms of molecular interactions to this kind of steroid activity.
Assuntos
Anabolizantes/química , Anabolizantes/farmacologia , Androgênios/química , Androgênios/farmacologia , Di-Hidrotestosterona/análogos & derivados , Modelos Químicos , Testosterona/análogos & derivados , Algoritmos , Androgênios/genética , Análise por Conglomerados , Simulação por Computador , Humanos , Masculino , Relação Quantitativa Estrutura-Atividade , Testosterona/genéticaRESUMO
The great cost associated with the development of new anabolic-androgenic steroid (AASs) makes necessary the development of computational methods that shorten the drug discovery pipeline. Toward this end, quantum, and physicochemical molecular descriptors, plus linear discriminant analysis (LDA) were used to analyze the anabolic/androgenic activity of structurally diverse steroids and to discover novel AASs, as well as also to give a structural interpretation of their anabolic-androgenic ratio (AAR). The obtained models are able to correctly classify 91.67% (86.27%) of the AASs in the training (test) sets, respectively. The results of predictions on the 10% full-out cross-validation test also evidence the robustness of the obtained model. Moreover, these classification functions are applied to an "in house" library of chemicals, to find novel AASs. Two new AASs are synthesized and tested for in vivo activity. Although both AASs are less active than some commercially AASs, this result leaves a door open to a virtual variational study of the structure of the two compounds, to improve their biological activity. The LDA-assisted QSAR models presented here, could significantly reduce the number of synthesized and tested AASs, as well as could increase the chance of finding new chemical entities with higher AAR.