Prediction of Ion Channel Activity Using Binary Kernel Discrimination

Willett, Peter; Wilton, David J.; Hartzoulakis, Basil; Tang, Rachel; Ford, John K. B.; Madge, David J.

doi:10.1021/ci700087v

Cited by 28 publications

(20 citation statements)

References 38 publications

(58 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Such models are based on frequency distributions of similarity values that are fused using integration over regions defined by the particular fusion rule. Typically, the use of binary kernel discrimination (BKD) for identifying potential active compounds in lead-discovery programs is superior to methods based on similarity searching and substructural analysis but inferior to a support vector machine [11, 13, 14, 17, 18]. New methods for ligand-based VS use data fusion and machine learning (ML) to enhance the effectiveness of identification of potential actives over typical similarity searching [19] using a single bioactive reference.…”

Section: Introductionmentioning

confidence: 99%

Brainstorming: weighted voting prediction of inhibitors for protein targets

Plewczynski

2010

J Mol Model

View full text Add to dashboard Cite

The “Brainstorming” approach presented in this paper is a weighted voting method that can improve the quality of predictions generated by several machine learning (ML) methods. First, an ensemble of heterogeneous ML algorithms is trained on available experimental data, then all solutions are gathered and a consensus is built between them. The final prediction is performed using a voting procedure, whereby the vote of each method is weighted according to a quality coefficient calculated using multivariable linear regression (MLR). The MLR optimization procedure is very fast, therefore no additional computational cost is introduced by using this jury approach. Here, brainstorming is applied to selecting actives from large collections of compounds relating to five diverse biological targets of medicinal interest, namely HIV-reverse transcriptase, cyclooxygenase-2, dihydrofolate reductase, estrogen receptor, and thrombin. The MDL Drug Data Report (MDDR) database was used for selecting known inhibitors for these protein targets, and experimental data was then used to train a set of machine learning methods. The benchmark dataset (available at http://bio.icm.edu.pl/∼darman/chemoinfo/benchmark.tar.gz) can be used for further testing of various clustering and machine learning methods when predicting the biological activity of compounds. Depending on the protein target, the overall recall value is raised by at least 20% in comparison to any single machine learning method (including ensemble methods like random forest) and unweighted simple majority voting procedures.

show abstract

Section: Introductionmentioning

confidence: 99%

Brainstorming: weighted voting prediction of inhibitors for protein targets

Plewczynski

2010

J Mol Model

View full text Add to dashboard Cite

show abstract

“…In LBVS, Bayesian methods have been widely applied over the past years, especially naive Bayesian classifiers [10], given their versatility, efficiency and ease of use. Binary kernel discrimination represents another Bayesian methodology adapted for LBVS [11], which utilizes fingerprints as descriptors. Furthermore, several extensions of Bayesian classification have been introduced and combinations with other approaches.…”

Section: Bayesian Methodsmentioning

confidence: 99%

Machine learning and similarity-based virtual screening techniques

Bajorath

2013

In Silico Drug Discovery and Design

View full text Add to dashboard Cite

Ligand-based virtual compound screening utilizes molecular similarity information. On the basis of calculated molecular similarity values, compounds are predicted to have a biological activity similar to known reference molecules. Approaches for ligand-based virtual screening can essentially be divided into similarity search and compound classification methods. For compound classification and database ranking, machine-learning approaches have become increasingly popular. A brief description of the ligand-based virtual screening field is provided and selected recent developments are discussed. Ligand-based virtual screening 136 Recent developments 139 Conclusion 144

show abstract

“…This publication started life in 1961 as the Journal of Chemical Documentation , at a time when the principal focus of the subject was the (printed) chemical literature. It changed its name to the Journal of Chemical Information and Computer Sciences in 1975 to reflect the by‐then central role of computer methods in the processing of chemical information, and adopted its current title as recently as 2005 to reflect the substantial changes that have taken place in the discipline over the past few years; the effects of these changes in bibliometric terms have been discussed elsewhere (Onodera, 2001; Willett, 2007). Apart from the Journal of Chemical Information and Modeling , the other most important journals for chemoinformatics material are the Journal of Computer‐Aided Molecular Design , Journal of Molecular Graphics and Modelling , and QSAR and Combinatorial Science , although an increasing number of journals across the chemical and life sciences now include chemoinformatics papers (Willett, 2007).…”

Section: The Emergence Of Chemoinformaticsmentioning

confidence: 99%