The 3D structure-based scorer is supported by the fragment-based (Fragment Growing and Fragment Linking) generators, but not the Fine-Tuning generator.
It can be combined with a 3D ligand-based scorer: for example, if the protein binding pocket is big and you want to avoid generating large compounds, a 3D ligand-based scorer will ensure that your generated molecules are close in shape and size to your reference compound.
To see a detailed walk-through demonstrating how you can set up a generator that uses 3D structure-based scores to optimize generated molecules, please have a look at the following use-case: 9: Growing around a fragment using 3D structure-based design.
1. Creation of the generator
NOTE: the reference molecule or initial fragments used in a generator should not contain charged atoms. Protonation is performed directly inside Makya.
To run a generator with structure-based parameters previously defined in the "Structure-Based Settings" tab:
- Create a new (Fragment Growing or Fragment Linking) generator,
- Enter all the constraints you wish to use in addition to your structure-based scorer (ex: QSAR, building blocks constraints, 3D ligand-based parameters...),
- Click on the "3D Structure based" tab,
- Select the 3D structure-based configuration you wish to use to guide generation.
A 3D visualization of the reference ligand and target protein pocket appears.
2. Selection of the anchor
If you click on the “Anchor” tab, you can set an anchor to be used during docking.
The anchor is a sub-unit of the reference molecule used as a starting point for 3D alignment. You can enter it as a SMILES or use the sketcher to draw it.
Once your anchor is set, click on Save.
IMPORTANT:
- This sub-unit has to be present in all the generated molecules for the 3D alignment to be done. Please make sure that your exit vector or building block constraints contain this part.
- The anchor should contain a minimum of 5 and a maximum of 25 heavy atoms.
3. Set-up of extra Scoring parameters
If you click on the "Scoring parameters" tab, you can customise the thresholds you wish to apply to the 3D structure-based scores.
During generation in Makya, the generative algorithm learns to design optimised molecules that fit your objectives. In the "Scoring parameters" tab, you can choose the values that the algorithm will use to assess if the molecules are good, or can be improved.
Default thresholds for good molecules are:
- Docking Score: < -10 for AutoDock Vina (on a range from -50 to 0), < -20 for rDock (on a range from -50 to 0)
- Contact Score: > 0.7 (on a range from 0 to 1)
Generation process and runtime
Below is a diagram indicating the different steps of a Makya generation using structure-based parameters and the approximate time spans associated to each step.
The longest process, with the highest impact on the total runtime, is the docking of generated molecules (several minutes per result). Docking in Makya is run in parallel, on up to several thousands molecules at a time.