These tools will no longer be maintained as of December 31, 2024. Archived website can be found here. PubMed4Hh GitHub repository can be found here. Contact NLM Customer Service if you have questions.


Pubmed for Handhelds

PUBMED FOR HANDHELDS

Search MEDLINE/PubMed

  • Title: Consensus adaptation of fields for molecular comparison (AFMoC) models incorporate ligand and receptor conformational variability into tailor-made scoring functions.
    Author: Breu B, Silber K, Gohlke H.
    Journal: J Chem Inf Model; 2007; 47(6):2383-400. PubMed ID: 17958410.
    Abstract:
    Taking into account dynamical behavior and/or structural inaccuracies of receptor-ligand systems becomes increasingly important in structure-based drug design. Here, we describe the development of consensus Adaptation of Fields for Molecular Comparison (AFMoC) (abbreviated as AFMoCcon) models that account for multiple ligand conformations in an ensemble of protein configurations. Ligand and receptor conformational variability is considered in a "reverse", protein-based CoMFA-type approach that results in a tailor-made scoring function. As an extension to the current AFMoC approach, AFMoCcon applies partial-least-squares regression considering multimode binding and a variable influence on the model-based region selection to an extended descriptor matrix. The approach was validated on a dataset of 79 structurally diverse thrombin inhibitors, aligned either in an experimentally determined thrombin structure or superimpositions of three structurally diverse thrombin structures derived by homology modeling. Initially, robust AFMoC models could be obtained for the experimental (q2 = 0.57) and one of the modeled protein structures (q2 = 0.61). However, no relationship between the quality of the homology model and the AFMoC model could be observed, rendering the a priori choice of a single receptor structure a difficult task. Convincingly, a consensus AFMoC model based on the newly developed approach circumvents this problem and shows a comparable internal and external predictivity (q2 = 0.61) like the best model derived from conventional AFMoC. As further advantages, (i) the influence of the single receptor structure-ligand alignment (RSLA) on the AFMoCcon model can be determined, (ii) there is no principal limitation regarding the number of different RSLA considered, and (iii) AFMoCcon models can be interpreted in terms of contour plots that aid in proposing variations of the ligand structure to improve binding. We expect the AFMoCcon approach also to be valuable in those cases where multiple experimental receptor conformations are given.
    [Abstract] [Full Text] [Related] [New Search]