Several encouraging pre‐clinical results highlight the melanin‐concentrating hormone receptor 1 (MCHR1) as promising target for anti‐obesity drug development. Currently however, experimentally resolved structures of MCHR1 are not available, which complicates rational drug design campaigns. In this study, we aimed at developing accurate, homologymodel‐based 3D pharmacophores against MCHR1. We show that traditional approaches involving docking of known active small molecules are hindered by the flexibility of binding pocket residues. Instead, we derived three‐dimensional pharmacophores from molecular dynamics simulations by employing our novel open‐source software PyRod. In a retrospective evaluation, the generated 3D pharmacophores were highly predictive returning up to 35 % of active molecules and showing an early enrichment (EF1) of up to 27.6. Furthermore, PyRod pharmacophores demonstrate higher sensitivity than ligand‐based pharmacophores and deliver structural insights, which are key to rational lead optimization.
