The question of open .dlg in ADT

[Degax]

When I use AutoDockTools to open a .dlg file generated by AutoDock-GPU, if any energy value (in any of the docking cycles) in the .dlg file is greater than 100, an error message appears: “invalid literal for float (): kcal/mol.” I’m not sure whether this issue is related to my CUDA configuration or if it’s a problem in the original AutoDock-GPU code.
For example, when the line reads “Final Total Internal Energy =+110.92 kcal/mol”, the error mentioned above appears when I try to open the file.

[diogomart]

Looks like AutoDockTools expects a space between the equal sign and the number. I suggest using meeko, ringtail, and pymol, see https://meeko.readthedocs.io

[Degax]

Looks like AutoDockTools expects a space between the equal sign and the number. I suggest using meeko, ringtail, and pymol, see https://meeko.readthedocs.io

First of all, I'm so glad that you can answer my question. Secondly, I have a question, please. If I don't open my .dlg file with adt and analyze it, I can't visualize this file (for example, use meeko to open it). I would like to ask if meeko can be visualized like adt?

Thank you very much.

[diogomart]

No, Meeko has no graphical interface, just command line scripts and Python bindings. Meeko's mk_export.py writes an SDF file that you can open with Pymol, ChimeraX, VMD, etc, as long as the input PDBQT for docking were also written be Meeko.

[Degax]

• [ ]

No, Meeko has no graphical interface, just command line scripts and Python bindings. Meeko's mk_export.py writes an SDF file that you can open with Pymol, ChimeraX, VMD, etc, as long as the input PDBQT for docking were also written be Meeko.

Thank you very much, sir!

Finally, I have one more question to ask. When I use AutoDock 4.0's de-docking, when the obtained dlg file is to analyze, a value of inhib_constant will appear. But I didn't have it when I used AutoDock-GPU. Excuse me, how can I get this value?

[diogomart]

Docking scores have big errors, converting them to binding constants can lead people to trust them too much. In any case, this may help: https://www.researchgate.net/post/Is-autodock-vina-binding-affinity-related-to-activation-energy

[Degax]

Docking scores have big errors, converting them to binding constants can lead people to trust them too much. In any case, this may help: https://www.researchgate.net/post/Is-autodock-vina-binding-affinity-related-to-activation-energy

Thank you very much, sir.

I still have a question. Before performing docking, I need to generate the corresponding map files. However, if I’m working with the same receptor but different ligands, I’m not quite sure how to efficiently generate all the necessary map files.

For example, when preparing for ligand A, AutoGrid might only generate NA.maps and C.maps, but then ligand B might require additional maps (like Cl.maps or O.maps) that weren’t created earlier. In this case, how can I quickly generate all the required maps for all ligands, without having to run AutoGrid separately for each one? Doing it one by one for every ligand is really time-consuming…

I’d really appreciate any suggestions or best practices. Thanks a lot in advance!

[diogomart]

Use the -g flag in meeko's mk_prepare_receptor.py
https://meeko.readthedocs.io/en/release-doc/rec_cli_options.html#cmdoption-g

It writes a gpf file with all ligand types