In Makya, you can use embedded docking to guide molecule generation by setting up 3D structure-based parameters. The 3D structure-based scores, such as Docking score and Contact Score will be optimized along with other specified scores (such as QSAR models), which means Makya will optimize the binding mode of the ligands toward the protein target in 3D.
Preparation and upload of the input files
To set up 3D Structure-based parameters in Makya, you will need to upload files that have been prepared beforehand. The preparation process is described in the following article:
Preparation of the Input Files
The files that have been prepared are uploaded into Makya, as described here:
Upload and Auto-preparation of 3D Files
Creation of new structure-based parameters
Once the files have been uploaded, click on the “3D Structure-Based” tab. This is where all your structure-based settings are saved in Makya.
Then create New Docking Parameters by clicking on the + logo.
1. Set up
A new window will appear and prompt you to enter a Name for your new structure-based parameters, choose the associated docking software, and select the docking files. Once it is done, click on Next.
2. Set your preferred protein-ligand interactions
You cannot use 3D SB parameters without selecting any interaction. It is necessary to have at least one protein ligand-interaction whose weight is ≥ 0. See below how to set up interaction weights.
Nomenclature of protein-ligand interactions
Once the files are selected, our in-house interaction detection software computes the interactions between the reference ligand and the protein atoms. For more details on the interactions we can detect, see the Advanced Information section at the end of this article.
A 3D visualization window appears, as well as a list of the detected interactions.
The labels for the atoms involved in a specific interaction are separated by a “+” sign. The first letter corresponds to the atom name (C for Carbon, H for Hydrogen, S for Sulfur…), while the second letter indicates the position of the atom on the side chains of the protein. Lastly, when a number is present in the atom label, it allows us to distinguish identical atoms at the same position.
Addition of new interactions
Sometimes, an interaction that is deemed important for compound activity can be absent from the list of detected interactions, because it was not found between the reference ligand and the protein. To generate compounds which form this new interaction, add the interaction by clicking on Interactions, then Add new, and select the type of contact, protein residue and protein atoms.
Setting the interaction weights
Once all relevant interactions have been selected and added, we can move on to the next step which is assigning weights to the interactions, using the sliding cursors.
Each interaction is assigned a respective weight. This weight represents how important the interaction is compared to other interactions of the same type. It is used by the Contact Score to compute a ratio between the number of interactions detected in a specific pose, versus the number of interactions specified by the user. What matters is the relative weights of all interactions whose weight is not 0; thus, tweaking the weight of a given interaction does not significantly impact the result.
Users should focus on putting weights on the most important interactions. Several methods can help determine which weights to prioritize:
- using a single PDB file with a co-crystallised ligand;
- using multiple PDB files with distinct ligands: the frequency of observed interactions gives a good indication of their relative importance;
- using Molecular Dynamics simulations: the frequency of observed interactions gives a good indication of their relative importance;
- using the user’s expert knowledge to manually tune them.
In the visualization, you will notice spheres on certain atoms that grow or shrink as you change interaction weights. These spheres represent the weight of each interaction. When an interaction involves a ring, spheres will appear on all necessary atoms. Dash lines are colored by interaction type; when hovering over the dash lines, you can see additional information related to the interaction types and involved atoms.
3. Set the exhaustiveness level (for AutoDock Vina only)
If the docking software is Vina, you can set the exhaustiveness to low, medium (recommended) or high. Vina exhaustiveness refers to the amount of computational effort involved in the docking.
- With a high exhaustiveness, Vina will sample more ligand conformations and orientations within the binding site, allowing for greater exploration, but also expensive and long docking.
- A low exhaustiveness will correspond to a fast, less exhaustive but also cheap docking.
The exhaustiveness parameter impacts the thoroughness, accuracy, and importantly, the computational resources and cost of docking.
Advanced information
Detection of interactions by our in-house technology
Interactions are detected by our software using several metrics (distances, angles, dihedral angles) and cut-off values. Below is a summary table of the implemented interactions and used criteria.