Cannot get valid molecule from running sample.py for a long time

[simmed00]

I read your paper which looks great.
I tried to run according to the instruction, sample.py, using the sample.yml and the two model ckpts downloaded from google drive. I waited for a long time (like half an hour) , and it still cannot produce a valid molecule, it shows [Pool] Queue 300 | Finished 0 | Failed 55.

I tried to debug, I found that most generated graph does not pass the if statement "if data_next.status == STATUS_FINISHED:".
Even for those few that can pass the if statement, it failed during "rdmol = reconstruct_from_generated(data_next)", which throws exception shown as "Ignoring, because reconstruction error encountered."

[luost26]

Hi, can you provide the index of the failed test data?

[simmed00]

Thank you for your prompt reply. I used the default index=0, I tried index=1, but the situation is same.
I run step and found that the reconstruction failed at line "rd_mol = convert_ob_mol_to_rd_mol(mol)"

[simmed00]

I further check downstream, it failed at "Chem.SanitizeMol(rd_mol,Chem.SANITIZE_ALL^Chem.SANITIZE_KEKULIZE)"

[luost26]

Thank you for your information. I am looking into it.

[jwelliavSG]

Hi, I am having the exact same issue - is this simply because the model takes a long time to sample valid molecules. I am getting 25 valid molecules after about 12 hours. 6590 were discounted either because they were duplicates or invalid.

[jwelliavSG]

Thanks in advance and also great paper ! Learnt a lot from it.

[tdby]

Hi, I am meeting the similar issue.
I have run the sample.py a few days,many samples_%d.pt are created,but it always shows [Pool] Queue 300 | Finished 0 | Failed 0.
When I debug it,I found that no one can pass the statement “if data_next.status == STATUS_FINISHED:”.
I tried to monitor the data“y_frontier” from the function “get_next” in sample.py, the frontiers always exist and status can not be “finished”.
Could you tell how can I solve it? Thank you!

[luost26]

Hi, I am meeting the similar issue. I have run the sample.py a few days,many samples_%d.pt are created,but it always shows [Pool] Queue 300 | Finished 0 | Failed 0. When I debug it,I found that no one can pass the statement “if data_next.status == STATUS_FINISHED:”. I tried to monitor the data“y_frontier” from the function “get_next” in sample.py, the frontiers always exist and status can not be “finished”. Could you tell how can I solve it? Thank you!

Can you try another random seed?

Btw, please check out our latest work on SBDD (accepted to ICML 2022) here: https://github.com/pengxingang/Pocket2Mol
It is significantly faster and generates molecule with higher quality.

[tdby]

Thanks for your reply.
I have tried several seeds including the example “4yhj.pdb” in sample_for_pdb.py, and some different data_id from “test_list.tsv” in sample.py.The result is the same. I think that is not my problem.

[luost26]

Thanks for your reply. I have tried several seeds including the example “4yhj.pdb” in sample_for_pdb.py, and some different data_id from “test_list.tsv” in sample.py.The result is the same. I think that is not my problem.

What's the version of RDKit and OpenBabel in your environment ?

[tdby]

RDkit==2022.03.3, and OpenBabel==3.1.1
Indeed it's different from your standard version.
Thanks for your advice,I will change it and try again.

[luost26]

RDkit==2022.03.3, and OpenBabel==3.1.1 Indeed it's different from your standard version. Thanks for your advice,I will change it and try again.

Thanks!
Again, I strongly recommend you to checkout our latest work here: https://github.com/pengxingang/Pocket2Mol

[tdby]

I will read it carefully.You and your team are very helpful for me.

[HaotianZhangAI4Science]

I thought it was caused by the the openbabel under windows