Combined Naïve Bayesian, Chemical Fingerprints and Molecular Docking Classifiers to Model and Predict Androgen Receptor Binding Data for Environmentally- and Health-Sensitive Substances.

Alfonso T García-Sosa,Uko Maran

doi:10.3390/ijms22136695

Alfonso T García-Sosa, Uko Maran

Open Access

https://doi.org/10.3390/ijms22136695

Copy DOI

Abstract

Many chemicals that enter the environment, food chain, and the human body can disrupt androgen-dependent pathways and mimic hormones and therefore, may be responsible for multiple diseases from reproductive to tumor. Thus, modeling and predicting androgen receptor activity is an important area of research. The aim of the current study was to find a method or combination of methods to predict compounds that can bind to and/or disrupt the androgen receptor, and thereby guide decision making and further analysis. A stepwise procedure proceeded from analysis of protein structures from human, chimp, and rat, followed by docking and subsequent ligand, and statistics based techniques that improved classification gradually. The best methods used multivariate logistic regression of combinations of chimpanzee protein structural docking scores, extended connectivity fingerprints, and naïve Bayesians of known binders and non-binders. Combination or consensus methods included data from a variety of procedures to improve the final model accuracy.

Highlights

Concern has been rising due to the fact that many environmental factors can modulate the androgen receptor (AR) pathway: agricultural and industrial chemicals, pharmacological drugs and chemotherapeutics, aging, hyperthermia, and chronic infection, alcohol, tobacco, and other drugs [1,2,3,4]
Different procedures applied to the training set and their sequential relationship are schematically viewed in Figure 3, where horizontal levels from top to bottom show the complexity of combinations and path towards the improvement of the Acc. and Matthews correlation coefficient (MCC) values
∗avgDInact + 3.584∗ PAct_dockChimp – 8.594∗PInact. This final model includes five descriptors: the docking score for the compound with chimp protein (ChimpDockScore), the average of the distances to the known active chemicals, the average of the distances to the known inactive chemicals, the Bayesian probability value for the compound according to the distribution of known actives compounds towards the chimp protein (PAct_dockChimp ), and the Bayesian probability value for the compound according to the distribution of known inactives compounds towards the chimp protein (PInact )

Summary

Introduction

Concern has been rising due to the fact that many environmental factors can modulate the androgen receptor (AR) pathway: agricultural and industrial chemicals, pharmacological drugs and chemotherapeutics, aging, hyperthermia, and chronic infection, alcohol, tobacco, and other drugs [1,2,3,4] These chemical agents enter the waterways, food chain, and affect other environments. Modulation of the AR has multiple biological effects on species in the environment, on health, and diseases These include important roles in the development and maintenance of reproductive [5], musculoskeletal [6,7], cardiovascular [8,9,10], immune [11,12], nervous [13,14,15], and hematopoietic [16,17] systems. AR binding compounds, or those that interfere with androgen-dependent path-

Objectives

Methods

Results

Discussion

Conclusion