Do I need to setup CCD and PDB PATH mirrors, if I just want to prepare .cdf and .sdf files from .pdb for use in RF3?

[tsamdin]

I'm trying to setup atomworks so I can prepare properly formatted .sdf/.cdf files for RF3 from .pdb files. I think I have the install setup correctly, but if I just want to set up a simple file conversion/formatting for RF3, do I necessarily need to download local copies of the CCD and PDB?

[nscorley]

No, it should be falling back to the Biotite CCD. What error are you seeing?

[tsamdin]

I'm trying to run the code provided in https://github.com/RosettaCommons/foundry/blob/production/models/rf3/README.md for creating custom cif files from an sdf input.

I was using a workflow where I took a pdb input into openbabel to get an sdf file, then put the sdf file into atomworks to get a cif file that RF3 likes. I found that simply exporting my pdb as a cif file from PyMol was not sufficient for RF3, some header information seemed to be missing?

Though I get the warning that the CCD and PDB mirror paths aren't set, I'm not sure if the default to Biotite is occuring and thus rasing the valence error, or if the problem is with how openbabel is gerneating my sdf input. Though I think it's the latter.

Environment variable CCD_MIRROR_PATH not set. Will not be able to use function requiring this variable. To set it you may:
(1) add the line 'export VAR_NAME=path/to/variable' to your .bashrc or .zshrc file
(2) set it in your current shell with 'export VAR_NAME=path/to/variable'
(3) write it to a .env file in the root of the atomworks.io repository
Environment variable PDB_MIRROR_PATH not set. Will not be able to use function requiring this variable. To set it you may:
(1) add the line 'export VAR_NAME=path/to/variable' to your .bashrc or .zshrc file
(2) set it in your current shell with 'export VAR_NAME=path/to/variable'
(3) write it to a .env file in the root of the atomworks.io repository
[08:50:32] Explicit valence for atom # 32 N, 4, is greater than permitted
[08:50:32] ERROR: Could not sanitize molecule ending on line 269
[08:50:32] ERROR: Explicit valence for atom # 32 N, 4, is greater than permitted
Traceback (most recent call last):
File "C:\Users...\RosettaFold_3\pdb2cif_aw.py", line 13, in
atom_array = atom_array_from_rdkit(sdf_to_rdkit(sdf_path))
^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users...\AppData\Roaming\Python\Python312\site-packages\atomworks\io\tools\rdkit.py", line 478, in sdf_to_rdkit
raise ValueError("Failed to parse molecule from SDF input") from None
ValueError: Failed to parse molecule from SDF input

I'm also exploring the tutorials in https://baker-laboratory.github.io/atomworks-dev/latest/auto_examples/load_and_visualize_structures.html - and trying to parse a simple script from the examples to prepare a cif/sdf/mol input that RF3 likes, when starting from a structure in a pdb format. I assume that this might be a better path forward for what I want to do?

[roccomoretti]

Explicit valence for atom # 32 N, 4, is greater than permitted

I'm guessing you have a positively charged nitrogen in your molecule which you don't have properly annotated as being positively charged. RDKit is rather picky about representations, and will fail if things aren't properly annotated.

A decent molecular editor should allow you to edit the charges on the atom. (e.g. with Avogadro you can go View>Properties>Atom Properties and edit the value in the table.) Or there may be a way to auto-detect charges. (e.g. with OpenBabel the "add hydrogens for pH" feature should fix charges on quaternary amines) Whatever way you do it, you'll want to make sure there's an M CHG line in your SDF which annotates the proper charge on your atom. (Probably will be something like M CHG 1 32 1, indicating one charge, on atom 32, with value +1)

[tsamdin]

Thanks for the suggestion about OpenBabel! I ended up removing the hydrogens entirely from my pdb file, and converting to sdf in openbabel using -p 7.4 and --gen3d. Atomworks made it further in the conversion, but now I get the following error:

(1) add the line 'export VAR_NAME=path/to/variable' to your .bashrc or .zshrc file
(2) set it in your current shell with 'export VAR_NAME=path/to/variable'
(3) write it to a .env file in the root of the atomworks.io repository
Environment variable PDB_MIRROR_PATH not set. Will not be able to use function requiring this variable. To set it you may:
(1) add the line 'export VAR_NAME=path/to/variable' to your .bashrc or .zshrc file
(2) set it in your current shell with 'export VAR_NAME=path/to/variable'
(3) write it to a .env file in the root of the atomworks.io repository

Failed to build _build_entity_poly: 'AtomArray' object has no attribute 'chain_iid'

Do I need to modify the AtomArray to create a chian_iid, or should my sdf file have it preserved before going in?

[nscorley]

Any chance you can share the input file that you're using and how you're running inference?