Induced-fit docking plus free energy perturbation brings structure-based design to promiscuous ADMET anti-targets
- Combines induced-fit docking with free energy perturbation to model flexible, promiscuous anti-target proteins: CYP3A4, CYP2D6, hERG and PXR.
- Imposing one shared binding pose across a congeneric series lets the workflow sample many receptor conformations efficiently.
- Evaluated retrospectively on public datasets and applied prospectively in ongoing drug discovery programs, supporting lead optimization away from off-target liabilities.
💡 Points Based on the abstract
The consensus-pose assumption is both the trick and the risk: for hERG and CYP3A4, ligands in one series can flip modes, and a forced shared pose will quietly bias the ranking. I'd want to see a no-FEP baseline (ligand-based QSAR or simple docking) on the same sets, plus the cases where the pose failed. Prospective use is encouraging, but hit rate is unreported here.
📄 Paper · Journal of Chemical Theory and Computation (ACS) ↗ Accelerating Lead Optimization Away from Anti-Targets: Navigating the Structural Complexity of Promiscuous ADMET Proteins