Drug Discovery Using Support Vector Machines. The Case Studies of Drug-likeness, Agrochemical-likeness, and Enzyme Inhibition Predictions

Zernov, Vladimir V.; Balakin, Konstantin V.; Ivaschenko, Andrey A.; Savchuk, Nikolay P.; Pletnev, I. V.

doi:10.1021/ci0340916

Cited by 193 publications

(116 citation statements)

References 20 publications

Supporting

Mentioning

114

Contrasting

Unclassified

Order By: Relevance

“…We calculated the EF′ value for the retrieval of the top-ranked 18 of the 25 known actives (72% of actives) in a given test set, close to the 70% chosen by Halgren et (11) al. 32 The advantage of this EF′ statistic is that if two different sets of molecules' SVM activity scores give the same ranking for the 18th active, the modified enrichment factor will yield a higher value for the set of scores that rank the other 17 actives higher in one list than in the other; the conventional enrichment factor (eq 10) would not make this distinction.…”

Section: Methodsmentioning

confidence: 99%

“…Several studies have shown the SVM to be among the best methods for correctly classifying molecules. [8][9][10][11] A standard application of the SVM algorithm involves defining two classes of objects, determining a set of numbers that characterize each object, and using the SVM algorithm to calculate a classification model for the objects. After this training step, the SVM model is used to classify other objects.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Virtual Screening of Molecular Databases Using a Support Vector Machine.

Jorissen

Gilson

2005

ChemInform

View full text Add to dashboard Cite

The Support Vector Machine (SVM) is an algorithm that derives a model used for the classification of data into two categories and which has good generalization properties. This study applies the SVM algorithm to the problem of virtual screening for molecules with a desired activity. In contrast to typical applications of the SVM, we emphasize not classification but enrichment of actives by using a modified version of the standard SVM function to rank molecules. The method employs a simple and novel criterion for picking molecular descriptors and uses cross-validation to select SVM parameters. The resulting method is more effective at enriching for active compounds with novel chemistries than binary fingerprint-based methods such as binary kernel discrimination.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Virtual Screening of Molecular Databases Using a Support Vector Machine.

Jorissen

Gilson

2005

ChemInform

View full text Add to dashboard Cite

show abstract

“…[1][2][3][4][5][6][7] These descriptors were initially developed for the construction of quantitative structure-activity relationship (QSAR) and quantitative structure-property relationship (QSPR) of structurally related compounds. 8 They have been extensively used for the statistical-learning-based prediction of pharmacodynamic, pharmacokinetic, and toxicological properties of chemical agents including drug-likeness, [9][10][11] blood-brain barrier penetration, 12,13 human intestinal absorption, 4 drug-receptor binding, [14][15][16] drug metabolism, 17 cellular membrane partitioning, 18 chemical space navigation, 19 and antibacterial activity. 20,21 Some of these molecular descriptors are developed for the study of a particular type of properties of a group of structurally related chemical agents.…”

Section: Introductionmentioning

confidence: 99%

Effect of Molecular Descriptor Feature Selection in Support Vector Machine Classification of Pharmacokinetic and Toxicological Properties of Chemical Agents

Yang

Yap

et al. 2004

J. Chem. Inf. Comput. Sci.

155

182

View full text Add to dashboard Cite

Statistical-learning methods have been developed for facilitating the prediction of pharmacokinetic and toxicological properties of chemical agents. These methods employ a variety of molecular descriptors to characterize structural and physicochemical properties of molecules. Some of these descriptors are specifically designed for the study of a particular type of properties or agents, and their use for other properties or agents might generate noise and affect the prediction accuracy of a statistical learning system. This work examines to what extent the reduction of this noise can improve the prediction accuracy of a statistical learning system. A feature selection method, recursive feature elimination (RFE), is used to automatically select molecular descriptors for support vector machines (SVM) prediction of P-glycoprotein substrates (P-gp), human intestinal absorption of molecules (HIA), and agents that cause torsades de pointes (TdP), a rare but serious side effect. RFE significantly reduces the number of descriptors for each of these properties thereby increasing the computational speed for their classification. The SVM prediction accuracies of P-gp and HIA are substantially increased and that of TdP remains unchanged by RFE. These prediction accuracies are comparable to those of earlier studies derived from a selective set of descriptors. Our study suggests that molecular feature selection is useful for improving the speed and, in some cases, the accuracy of statistical learning methods for the prediction of pharmacokinetic and toxicological properties of chemical agents.

show abstract

“…Similar trends between databases containing drugs, i.e., the MACCS-I Drug Data Report (MDDR) as well as the Comprehensive Medicinal Chemistry (CMC) and compound collections comprising mainly nondrugs such as the Available Chemical Directory (ACD), were reported before by Oprea and later by Zheng et al 29,51 While Figure 3a shows a clear separation of the maximum peaks of the molecular weight of nondrugs and drugs, Zernov et al yielded a stronger overlap between both distributions that intersect at around 360 that is higher than our separating margin. 19 Almost complete overlap between substances from the World Drugs Index (WDI) and those from the ACD were found by Li et al 17 Another obvious visual separation is found for the molar refractivity (see Figure 3b). The separating margin obtained by tree B is 40, identical to that obtained from tree A.…”

Section: Resultsmentioning

confidence: 89%

“…7,[11][12][13][14][15][16][17] However, even support vector machines that are known to be among the most accurate methods for classification tasks have so far not provided fully satisfactory results. [17][18][19][20] This can be, at least in part, explained by the implicit difficulties that arise from the data set.…”

Section: Introductionmentioning

confidence: 99%

Gradual in Silico Filtering for Druglike Substances

Schneider

Jäckels

Andrès

et al. 2008

J. Chem. Inf. Model.

View full text Add to dashboard Cite

The suitability of decision trees in comparison to support vector machines for the classification of chemical compounds into drugs and nondrugs was investigated. To account for the requirements upon screening virtual compound libraries, schemes for successive filtering steps with gradual increasing computational cost are outlined. The obtained prediction accuracy was similar between decision trees and support vector machine approaches for the applied compound data sets. By using rapidly computable variables such as druglikeness indices, XlogP, and the molar refractivity, at least 39% of the nondrugs can be filtered out, while retaining more than 83% of the actual drugs. Computationally more demanding descriptors such as specific substructure queries and quantum chemically derived variables can be postponed to subsequent classification schemes for the reduced set of compounds, whereby up to 92% of the nondrugs can be sorted out without loosing considerably more drugs. Using all available computed descriptors simultaneously in the first step did not yield significantly better results. Furthermore, the generated decision trees are used to derive guidelines for the design of druglike substances. The numerical margins found at the branching points suggest several criteria that separate drugs from nondrugs: a molecular weight higher than 230, a molar refractivity higher than 40, and the presence of one or more rings as well as one or more functional groups. Also reported are additionally required parameters to compute values for XlogP, SlogP, and the molar refractivity of boron and silicon containing compounds.

show abstract

Drug Discovery Using Support Vector Machines. The Case Studies of Drug-likeness, Agrochemical-likeness, and Enzyme Inhibition Predictions

Cited by 193 publications

References 20 publications

Virtual Screening of Molecular Databases Using a Support Vector Machine.

Virtual Screening of Molecular Databases Using a Support Vector Machine.

Effect of Molecular Descriptor Feature Selection in Support Vector Machine Classification of Pharmacokinetic and Toxicological Properties of Chemical Agents

Gradual in Silico Filtering for Druglike Substances

Contact Info

Product

Resources

About