Selecting Relevant Descriptors for Classification by Bayesian Estimates: A Comparison with Decision Trees and Support Vector Machines Approaches for Disparate Data Sets

Abstract

Classification algorithms suffer from the curse of dimensionality, which leads to overfitting, particularly if the problem is over-determined. Therefore it is of particular interest to identify the most relevant descriptors to reduce the complexity. We applied Bayesian estimates to model the probability distribution of descriptors values used for binary classification using n-fold cross-validation. As a measure for the discriminative power of the classifiers, the symmetric form of the Kullback-Leibler divergence of their probability distributions was computed. We found that the most relevant descriptors possess a Gaussian-like distribution of their values, show the largest divergences, and therefore appear most often in the cross-validation scenario. The results were compared to those of the LASSO feature selection method applied to multiple decision trees and support vector machine approaches for data sets of substrates and nonsubstrates of three Cytochrome P450 isoenzymes, which comprise strongly unbalanced compound distributions. In contrast to decision trees and support vector machines, the performance of Bayesian estimates is less affected by unbalanced data sets. This strategy reveals those descriptors that allow a simple linear separation of the classes, whereas the superior accuracy of decision trees and support vector machines can be attributed to nonlinear separation, which are in turn more prone to overfitting.

SwissADME: a free web tool to evaluate pharmacokinetics, drug-likeness and medicinal chemistry friendliness of small molecules

“…55) were built on severely unbalanced training sets and tested on clearly unbalanced external sets. As demonstrated by Carbon-Mangels et al 56. the relevance of machine-learning classification methods, and especially SVM, are negatively impacted by datasets with one significantly more populated class.…”

Selective Hypoxia‐Cytotoxin 7‐Fluoro‐2‐Aminophenazine 5,10‐Dioxide: Toward “Candidate‐to‐Drug” Stage in the Drug‐Development Pipeline

“…Non‐inhibition or ability to be substrate is labeled as (‐); [d] Renal organic cation transporter (OCT2) inhibition . Non‐inhibition is labeled as (‐); [e] Ability to be substrate or inhibit Cytochrome P450 isoenzymes (CYP450 2 C9, CYP450 2D6, CYP450 3 A4, CYP450 1 A2, and CYP 2 C19) . It is shown the predicted isoform‐CYP that the compound potentially inhibits; [f] Human ether‐a‐go‐go‐related gene (hERG) inhibition .…”

“…[28] Non-inhibition is labeled as (-); [e] Ability to be substrate or inhibit Cytochrome P450 isoenzymes (CYP450 2 C9, CYP450 2D6, CYP450 3 A4, CYP450 1 A2, and CYP 2 C19). [29] It is shown the predicted isoform-CYP that the compound potentially inhibits; [f] Human ether-a-go-go-related gene (hERG) inhibition. [30,31] Non-inhibition is labeled as (-); [g] In vitro mutagenicity according to Ames test.…”

In silico approach for the discovery of new PPARγ modulators among plant-derived polyphenols

“…An ADME analysis includes the analysis of various properties such as ability to penetrate blood–brain barrier, 43 capability of human intestinal absorption, 43 Caco-2 permeability, 44 and abilities to function as a P-glycoprotein (P-gp) substrate 45 and inhibitor, 46 , 47 renal organic cation transporter, 48 and cytochrome P450 substrate 49 and inhibitor. 50 …”