Following interface can be used to generate a quick visualization of results from druggability analysis.
Input files are :file:`prefix_heavyatoms.pdb` and other PDB files in :file:`prefix` folder.
- If you have outputted aligned trajectory in grid calculation step, you can select to load it too.
- Optionally, molecules present in VMD can be deleted, high resolution representations and a protein surface representation can be generated.
When results are loaded, you will see a representation similar to the following:
Each sphere corresponds to a probe binding spot. Spheres are colored according to their binding free energies. Red most sphere has the lowers binding free energy.
Binding free energies of probes can be found in the logfile:
.. literalinclude:: drugui_tutorial_files/sample.log
Logfile lists all probe binding spots, their binding free energies, and fractional contribution of different probe types to the hotspot.
Druggable sites are identified by clustering probe binding spots and merging them to identify subsets of binding spots that have a size similar to that of a drug-like molecule. After results load, you will see a list of molecules in :guilabel:`VMD Main` for each druggable site and solutions therein. You can toggle displayed molecules to see locations of different sites and solutions.
Figure shows the best solution for protein MDM2. The maximal achievable affinity (druggability index) for this solution is 0.3 nM or, in terms of free energy, it is -13 kcal/mol. You can find such information in the log file shown above.
Probe binding hotspots and protein structure shown above can be found among tutorial files. These results of course deserve a more detailed analysis, and some things that can be done include:
- looking into types of probes that contribute to a given binding spot and types of amino acid residue interacting with the binding spot
- visualizing trajectories (in which probes are wrapped) to see specific interactions and residence time of probes at a given binding spot
- comparing results from simulations in presence and absence of probes to see how binding site shape is affected by ligand binding
- looking into other structures of the target protein (ligand binding sites, crystal contacts, protein interfaces) to see whether observations in simulations are supported by interactions determined experimentally
Finally, you can visualize probe occupancy grids using VMD. Simply load :file:`.dx` files and create :guilabel:`isovolume` representations. An example is shown below for :file:`sample_IPRO.dx`, and other grid files. Mesh surfaces correspond to locations that were highly enriched with probes. Coloring is as follows:
- isopropanol: green (high enrichment at the binding site)
- acetate: red (enrichment at the surface, not that proetein has +5 net charge)
- isopropylamine: blue (few interaction spots)
- acetamide: magenta (not observed to interact with this protein much)
Note that values in occupancy grids is the count of central carbon atoms of probe molecules. Since the grid elements (voxels) are small (0.5A dimension), the occupancy numbers are small. You will need to adjust :guilabel:`Isovalue` value in VMD Representations window to make grid elements visible.
Similar representations can be generated for water or other atom type specific grids too.