Practical application of the Average Information Content Maximization (AIC-MAX) algorithm: selection of the most important structural features for serotonin receptor ligands

Warszycki, Dawid; Śmieja, Marek; Kafel, Rafał

doi:10.1007/s11030-017-9729-8

Cited by 3 publications

(3 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since the time complexity of the simplest logistic regression model is linear with respect to data dimension, the use of Pharmacoprint (39 973 bits) can be around 39 times slower than typical structural fingerprints, such as ECFP4 or Ext (1024 bits). To reduce the time processing of Pharmacoprint, one can consider the application of reduction algorithms, which allows one to not only save CPU time but also improve the results. ,, For this purpose, 100-dimensional representations generated by PCA, AE, and sAE were considered (although the number of 100 bits has been taken as an exemplary value, similar MCC values are obtained for the other cutoffs; see Figures S8–S10). An additional case which considers the number of principal components that explain 90% of the variance was also studied.…”

Section: Results and Discussionmentioning

confidence: 99%

Pharmacoprint: A Combination of a Pharmacophore Fingerprint and Artificial Intelligence as a Tool for Computer-Aided Drug Design

Warszycki

Struski

Śmieja

et al. 2021

J. Chem. Inf. Model.

Self Cite

View full text Add to dashboard Cite

Structural fingerprints and pharmacophore modeling are methodologies that have been used for at least 2 decades in various fields of cheminformatics, from similarity searching to machine learning (ML). Advances in in silico techniques consequently led to combining both these methodologies into a new approach known as the pharmacophore fingerprint. Herein, we propose a high-resolution, pharmacophore fingerprint called Pharmacoprint that encodes the presence, types, and relationships between pharmacophore features of a molecule. Pharmacoprint was evaluated in classification experiments by using ML algorithms (logistic regression, support vector machines, linear support vector machines, and neural networks) and outperformed other popular molecular fingerprints (i.e., ECFP4, Estate, MACCS, PubChem, Substructure, Klekota–Roth, CDK, Extended, and GraphOnly) and the ChemAxon pharmacophoric features fingerprint. Pharmacoprint consisted of 39 973 bits; several methods were applied for dimensionality reduction, and the best algorithm not only reduced the length of the bit string but also improved the efficiency of the ML tests. Further optimization allowed us to define the best parameter settings for using Pharmacoprint in discrimination tests and for maximizing statistical parameters. Finally, Pharmacoprint generated for three-dimensional (3D) structures with defined hydrogens as input data was applied to neural networks with a supervised autoencoder for selecting the most important bits and allowed us to maximize the Matthews correlation coefficient up to 0.962. The results show the potential of Pharmacoprint as a new, perspective tool for computer-aided drug design.

show abstract

Section: Results and Discussionmentioning

confidence: 99%

Pharmacoprint: A Combination of a Pharmacophore Fingerprint and Artificial Intelligence as a Tool for Computer-Aided Drug Design

Warszycki

Struski

Śmieja

et al. 2021

J. Chem. Inf. Model.

Self Cite

View full text Add to dashboard Cite

show abstract

“…Notably, each bit of fingerprint was considered as a single feature in our study; thus, the optimal feature set comprises hybrid fingerprints and descriptors after the feature reduction process. By removing irrelevant bits from the original intact fingerprint, a hybrid fingerprint can achieve increased prediction accuracy as well as reduced computational cost (Williams, 2006; Nisius and Bajorath, 2009, 2010; Singla et al, 2013; Smieja and Warszycki, 2016; Warszycki et al, 2017).…”

Section: Algorithms and Methodsmentioning

confidence: 99%

Development of Predictive Models for Identifying Potential S100A9 Inhibitors Based on Machine Learning Methods

et al. 2019

View full text Add to dashboard Cite

S100A9 is a potential therapeutic target for various disease including prostate cancer, colorectal cancer, and Alzheimer's disease. However, the sparsity of atomic level data, such as protein-protein interaction of S100A9 with RAGE, TLR4/MD2, or CD147 (EMMPRIN) hinders the rational drug design of S100A9 inhibitors. Herein we first report predictive models of S100A9 inhibitory effect by applying machine learning classifiers on 2D-molecular descriptors. The models were optimized through feature selectors as well as classifiers to produce the top eight random forest models with robust predictability and high cost-effectiveness. Notably, optimal feature sets were obtained after the reduction of 2,798 features into dozens of features with the chopping of fingerprint bits. Moreover, the high efficiency of compact feature sets allowed us to further screen a large-scale dataset (over 6,000,000 compounds) within a week. Through a consensus vote of the top models, 46 hits (hit rate = 0.000713%) were identified as potential S100A9 inhibitors. We expect that our models will facilitate the drug discovery process by providing high predictive power as well as cost-reduction ability and give insights into designing novel drugs targeting S100A9.

show abstract

“…Notably, each bit of fingerprint was considered as a single feature in our study, thus the optimal feature set comprises of hybrid fingerprints and descriptors after feature reduction process. By removing irrelevant bits from original intact fingerprint, hybrid fingerprint can achieve increased prediction accuracy, as well as reduced computational cost [51][52][53][54][55][56].…”

Section: [Figure1 Insert Here]mentioning

confidence: 99%

Development of Predictive Models for Identifying Potential S100A9 Inhibitors Based on Machine Learning Methods

Lee¹,

Kumar²,

Lee³

et al. 2019

Preprint

View full text Add to dashboard Cite

S100A9 is a potential therapeutic target for various disease including prostate cancer, colorectal cancer, and Alzheimer’s disease. However, the sparsity of atomic level data such as protein-protein interaction of S100A9 with MD2/TLR4/CD147 makes rational drug design of S100A9 inhibitors more challengeable. Herein we firstly report predictive models of S100A9 inhibitory effect by applying machine learning classifiers on 2D-molecular descriptors. The models were optimized through feature selectors as well as classifiers to produce the top eight random forest models with robust predictability as well as high cost-effectiveness. Notably, the optimal feature sets were obtained after the reduction of 2798 features into dozens of features with the chopping of fingerprint bits. In addition, the high efficiency of compact feature sets allowed us to further screen a large-scale dataset (over 6,000,000 compounds) within a week. Through the consensus vote of the top models, 46 hits (hit rate = 0.000713%) were identified as potential S100A9 inhibitors. We expect that our models will facilitate the drug discovery process by providing high predictive power as well as cost-reduction ability and give insights into the design of the novel drugs targeting S100A9.

show abstract

Practical application of the Average Information Content Maximization (AIC-MAX) algorithm: selection of the most important structural features for serotonin receptor ligands

Cited by 3 publications

References 32 publications

Pharmacoprint: A Combination of a Pharmacophore Fingerprint and Artificial Intelligence as a Tool for Computer-Aided Drug Design

Pharmacoprint: A Combination of a Pharmacophore Fingerprint and Artificial Intelligence as a Tool for Computer-Aided Drug Design

Development of Predictive Models for Identifying Potential S100A9 Inhibitors Based on Machine Learning Methods

Development of Predictive Models for Identifying Potential S100A9 Inhibitors Based on Machine Learning Methods

Contact Info

Product

Resources

About