Multi-Step Protocol for Automatic Evaluation of Docking Results Based on Machine Learning Methods—A Case Study of Serotonin Receptors 5-HT₆ and 5-HT₇

Abstract

Molecular docking, despite its undeniable usefulness in computer-aided drug design protocols and the increasing sophistication of tools used in the prediction of ligand-protein interaction energies, is still connected with a problem of effective results analysis. In this study, a novel protocol for the automatic evaluation of numerous docking results is presented, being a combination of Structural Interaction Fingerprints and Spectrophores descriptors, machine-learning techniques, and multi-step results analysis. Such an approach takes into consideration the performance of a particular learning algorithm (five machine learning methods were applied), the performance of the docking algorithm itself, the variety of conformations returned from the docking experiment, and the receptor structure (homology models were constructed on five different templates). Evaluation using compounds active toward 5-HT6 and 5-HT7 receptors, as well as additional analysis carried out for beta-2 adrenergic receptor ligands, proved that the methodology is a viable tool for supporting virtual screening protocols, enabling proper discrimination between active and inactive compounds.

N-(2-Hydroxyphenyl)-1-[3-(2-oxo-2,3-dihydro-1H- benzimidazol-1-yl)propyl]piperidine-4-Carboxamide (D2AAK4), a Multi-Target Ligand of Aminergic GPCRs, as a Potential Antipsychotic

“…In the case of serotonin 5-HT 7 receptor, first, we constructed a homology model of this protein. Our model is an addition to previously published homology models of serotonin 5-HT 7 receptor [46][47][48][49] and is in general accordance with them. Molecular interactions of D2AAK4 with the studied aminergic GPCRs are typical for ligands with a protonatable nitrogen atom interacting with the conserved Asp 3.32 as the main anchoring point [50].…”

Life beyond the Tanimoto coefficient: similarity measures for interaction fingerprints

“…for the discovery of GPCR (G-protein coupled receptor) ligands [15] or kinase inhibitors [16]. In more complex examples, they have been applied for interpreting activity landscapes [17], for training machine learning models [18], and for identifying covalently targetable cysteine residues in the human kinome [19]. Additionally, interaction fingerprints are applied to support large, specialized structural databases, such as GPCRdb (for GPCRs) [20], KLIFS (for kinases) [21, 22] or PDEstrian (for phosphodiesterases) [23].…”

Average Information Content Maximization—A New Approach for Fingerprint Hybridization and Reduction

“…The quality of the bits chosen by the AIC-Max algorithm was verified in a classification experiment conducted for the 5 underlying serotonin receptor ligands. As a classification method, a random forests technique [ 19 ] implemented in randomForest R package was used because it is known to be one of the state-of-the-art approaches in activity prediction [ 6 ]. The accuracy of classification was evaluated via Matthews Correlation Coefficient ( MCC ), the well-known validation measure, especially for imbalanced datasets.…”