Predicting a Binding Site (predict.py): TypeError: sequence item 0: expected str instance, numpy.float32 found

[Foly93]

OS: Ubuntu 22.04 mamba: 1.5.5 conda: 23.11.0

Hi,

I am trying to use your program to predict binding sites in a protein PDB. when I run

python ~/DeepPocket/predict.py -p 1A9N_frame_0.pdb -c ~/DeepPocket/first_model_fold1_best_test_auc_85001.pth.tar -s ~/DeepPocket/seg0_best_test_IOU_91.pth.tar -r 3

I get the following terminal message:

/home/luis/DeepPocket/rank_pockets.py:87: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.
  all_probs.append(F.softmax(output).detach().cpu())
Traceback (most recent call last):
  File "/home/luis/DeepPocket/predict.py", line 106, in <module>
    fout.write(''.join(confidence_types))
TypeError: sequence item 0: expected str instance, numpy.float32 found

I have tried to circumvent the TypeError by changing line 106 in predict.py from fout.write(''.join(confidence_types)) to fout.write(''.join(str(confidence_types))), but that only led to yet another more cryptic error after prompting the aforementioned python command:

Traceback (most recent call last):
  File "/home/luis/DeepPocket/predict.py", line 122, in <module>
    test(seg_model, seg_eptest, seg_gmaker,device,dx_name, args)
  File "/home/luis/DeepPocket/segment_pockets.py", line 145, in test
    output_pocket_pdb(dx_name+'_pocket'+str(count)+'.pdb',prot_prody,pred_aa)
  File "/home/luis/DeepPocket/segment_pockets.py", line 85, in output_pocket_pdb
    pocket=prot_prody.select(sel_str)
  File "/home/luis/miniforge3/envs/DeepPocketEnv/lib/python3.9/site-packages/prody/atomic/atomic.py", line 232, in select
    return SELECT.select(self, selstr, **kwargs)
  File "/home/luis/miniforge3/envs/DeepPocketEnv/lib/python3.9/site-packages/prody/atomic/select.py", line 894, in select
    indices = self.getIndices(atoms, selstr, **kwargs)
  File "/home/luis/miniforge3/envs/DeepPocketEnv/lib/python3.9/site-packages/prody/atomic/select.py", line 952, in getIndices
    torf = self.getBoolArray(atoms, selstr, **kwargs)
  File "/home/luis/miniforge3/envs/DeepPocketEnv/lib/python3.9/site-packages/prody/atomic/select.py", line 1003, in getBoolArray
    parser = self._getParser(selstr)
  File "/home/luis/miniforge3/envs/DeepPocketEnv/lib/python3.9/site-packages/prody/atomic/select.py", line 1102, in _getParser
    parser.enablePackrat()
  File "/home/luis/miniforge3/envs/DeepPocketEnv/lib/python3.9/site-packages/pyparsing/util.py", line 265, in _inner
    return fn(*args, **kwargs)
  File "/home/luis/miniforge3/envs/DeepPocketEnv/lib/python3.9/site-packages/pyparsing/core.py", line 1132, in enable_packrat
    ParserElement.packrat_cache = _FifoCache(cache_size_limit)  # type: ignore[assignment]
  File "/home/luis/miniforge3/envs/DeepPocketEnv/lib/python3.9/site-packages/pyparsing/util.py", line 105, in __init__
    keyring = [object()] * size
TypeError: can't multiply sequence by non-int of type 'Forward'

Do you have an idea how to fix these issues?

Cheers Foly

[RishalAggarwal]

thank you for the fix on the initial error, it seems like your second error seems to be related prody/pyparsing compatibility, maybe this would help: https://github.com/prody/ProDy/issues/1856

[Foly93]

Hi,

it does work now, with both these issues fixed.

Maybe it is helpful for future users to provide a YAML file for easy python environment setup. I attached my environment YAML file to this message, which creates (a hopefully working) environment via

mamba env create -n DeepPocketEnv -f DeepPocketEnv.yaml

or instead of mamba, conda and micromamba should also work.

Cheers, Luis

DeepPocketEnv.zip