Our approach is a ligand-based method where the most active compound will be used as a reference to find potential MCH antagonists. Our method partitions the reference structure in multiple fragments and compares each of the fragments against a library of the building blocks used to create the EnamineREAL library. The selection of the best building blocks is performed using the hydrophobic profile in 3D derived from QM-based descriptors. In order to align and compare the molecules, a low energy/bioactive conformer has to be generated for the reference fragments. This conformer is then superposed to multiple conformers of each building block to take into account molecular flexibility. The alignment of both structures is carried out by maximizing the similarity of their hydrophobic fields, which will be utilized for comparing the two structures. The hydrophobic field is derived from the fractional hydrophobic contributions within the quantum mechanical version of the Miertus−Scrocco−Tomasi (MST) continuum model, which provides a structure-agnostic descriptor that increases the diversity of the selected compounds. From the comparison of each reference fragment against the building block library, a list of prioritized building blocks is generated. Then, these building blocks are combined using the associated chemical reactions provided by Enamine to enumerate a library of candidate compounds; generating only structures that can be purchased from the library. If enough information of the binding mode of the known active compounds is gathered, more than one reference structure might be selected for screening to take into account the different binding modes.
Optionally a second step to refine the selected compounds might be performed. If most structures share the binding mode, 3D QSAR methods could be used to generate a model to predict the activity of the enumerated compounds. In order to do that a 3D overlay of the provided structures will be generated and the same hydrophobic profile will be used to generate a 3D QSAR model. Then, this model will be used to predict the activity of the enumerated compounds and refine the selection. Alternatively, if the 3D QSAR model quality is not good, selected compounds can also be docked to the Alphafold structure to refine the final compound selection. This second step will be more important in the Hit Optimization stage.
