Using a staged multi-objective optimization approach to find selective pharmacophore models

Abstract

It is often difficult to differentiate effectively between related G-protein coupled receptors and their subtypes when doing ligand-based drug design. GALAHAD uses a multi-objective scoring system to generate multiple alignments involving alternative trade-offs between the conflicting desires to minimize internal strain while maximizing pharmacophoric and steric (pharmacomorphic) concordance between ligands. The various overlays obtained can be associated with different subtypes by examination, even when the ligands available do not discriminate completely between receptors and when no specificity information has been used to bias the alignment process. This makes GALAHAD a potentially powerful tool for identifying discriminating models, as is illustrated here using a set of dopaminergic agonists that vary in their D1 vs. D2 receptor selectivity.

A Two-Step Target Binding and Selectivity Support Vector Machines Approach for Virtual Screening of Dopamine Receptor Subtype-Selective Ligands

“…These are consistent with a CoMSIA based analysis that suggests that D2 selectivity is determined by hydrogen bond acceptor (linked to H-bond acceptor) and donor (linked to HaaNH), hydrophobic (linked to HCsats, dssC, aasC), and electrostatic (linked to HaaNH, aaNH) factors [10]. These are also consistent with the conclusion from a pharmacophoric model that two hydrogen acceptors or one hydrogen acceptor plus one donor are critically important for D2 selectivity of some ligands [14].…”

“…The top-ranked D1 selective descriptors are number of O atoms, sum of Estate of atom type dssC, ssO and ssNH, graph-theoretical shape coefficient, and sum of H Estate of atom type HsNH2. These descriptors are consistent with the D1 selective features derived from a pharmacophoric model that includes positive nitrogens (linked to ssNH, HsNH2), hydrogen bond acceptor (linked to O, ssO) and donor (linked to ssNH, HsNH2) [14]. The top-ranked D2 selective descriptors are number of H-bond acceptor, sum of H Estate of atom types HaaNH and HCsats, and sum of Estate of atom type dssC, aasC and aaNH.…”

“…For instance, 3D-QSAR models have been developed for D2, D3 and D4 selective ligands respectively, achieving good prediction performances with R 2 and Q 2 values in the ranges of 0.89–0.97 and 0.58–0.84 respectively [10], [11], [12], [13]. A GALAHAD based selective pharmacophore model has been constructed for D1/D2 selective agents [14]. CoMFA and CoMSIA models have been developed for D2, D3 and D4 selective ligands [15].…”

Rational questing for inhibitors of endothelin converting enzyme-1 from Salvia miltiorrhiza by combining ligand- and structure-based virtual screening

“…22 They are adopted to assess reproductive fitness and select which candidates should survive to the next generation by making use of the Pareto rank for each individual model. 23 The GALAHAD models derived from the training set were compared according to Pareto ranking and a small value of SE and high values of SO and PhS were considered to be a good model. Before establishing the model, all of the compounds were prepared as follows: the structures were checked for bond orders, hydrogen atoms were added, and a minimization procedure was implemented using the MMFF94 force-field.…”

3D-QSAR (CoMFA and CoMSIA) and pharmacophore (GALAHAD) studies on the differential inhibition of aldose reductase by flavonoid compounds

“…This MO functions are employed for three different purposes: to assess reproductive fitness, to select which candidates should survive to the next generation, and to rank models after Cartesian alignment of their constituent ligand conformers. The three MO functions constitute a multi-objective triage (MOTriage) approach, which make use of the Pareto rank for each individual model [36].…”