Hi everyone I m trying to perform an AF test using the pipel nf-core #proteinfold

Hi everyone, I;m trying to perform an AF test usin...

Mauricio Morales

03/26/2025, 3:58 PM

Hi everyone, I;m trying to perform an AF test using the pipeline. I'm getting the following error:

Copy code

Command error:
  INFO:    Environment variable SINGULARITYENV_TMPDIR is set, but APPTAINERENV_TMPDIR is preferred
  INFO:    Environment variable SINGULARITYENV_NXF_TASK_WORKDIR is set, but APPTAINERENV_NXF_TASK_WORKDIR is preferred
  INFO:    Environment variable SINGULARITYENV_NXF_DEBUG is set, but APPTAINERENV_NXF_DEBUG is preferred
  E0326 15:47:21.546588 140614903998272 templates.py:837] Could not find CIFs in ./pdb_mmcif/mmcif_files
  Traceback (most recent call last):
    File "/app/alphafold/run_alphafold.py", line 570, in <module>
      app.run(main)
    File "/conda/lib/python3.10/site-packages/absl/app.py", line 312, in run
      _run_main(main, args)
    File "/conda/lib/python3.10/site-packages/absl/app.py", line 258, in _run_main
      sys.exit(main(argv))
    File "/app/alphafold/run_alphafold.py", line 467, in main
      template_featurizer = templates.HmmsearchHitFeaturizer(
    File "/app/alphafold/alphafold/data/templates.py", line 838, in __init__
      raise ValueError(f'Could not find CIFs in {self._mmcif_dir}')
  ValueError: Could not find CIFs in ./pdb_mmcif/mmcif_files

I followed the expected structure of the

pdb_mmcif

directory:

Copy code

├── pdb_mmcif
│   ├── mmcif_files
│   │   ├── 3wae.cif
│   │   ├── 3wt1.cif
│   │   ├── 3wtr.cif
│   │   ├── 3wy2.cif
│   │   ├── 3zud.cif
│   │   ├── 4bix.cif
│   │   ├── 4bzx.cif
│   │   ├── 4c1n.cif
│   │   ├── 4cej.cif
│   │   ├── 4chm.cif
│   │   ├── 4fzo.cif
...
└── obsolete.dat

Questions: • Should I unzip all the files? • I'm running the pipeline using

Singularity

, I checked different options and some comments report that could be an issue related on how

Singularity

use the paths. This looks more like an issue with

Singularity

rather that the structure of the

DBs

, am I correct? So far all the

DBs

are okay:

Copy code

bfd  mgnify  params  pdb70  pdb_mmcif  pdb_seqres  small_bfd  uniprot  uniref30  uniref90

Anyone had the same issue in the past? Any comment is more than welcome! Thank you everyone

Jose Espinosa-Carrasco

03/26/2025, 4:29 PM

Hi! Are you using the last

dev

version of the pipeline? this was fixed yesterday on this PR

Mauricio Morales

03/26/2025, 5:16 PM

Thank you for the quick response, yes I'm using the dev version, I will check if I can solve the issue with the link that you provided

Mauricio Morales

03/26/2025, 8:52 PM

Hi again, I followed the recommendation pointed in the link:

you will need to update this line in the
schema
to
"default": "null/pdb_mmcif/mmcif_files/*

What I changed? In the

nextflow_schema.json

I edit the following lines:

Copy code

"pdb_mmcif_path": {
"type": "string",
"description": "Path to the PDB mmCIF database",
"fa_icon": "fas fa-folder-open"
},

adding the following line at the end:

Copy code

"default": "null/pdb_mmcif/*"
},

In this occasion the pipeline ran for 16 minutes instead of the initial 7 minutes that last all the previous one with the same error. What do you think? Did I forget something? Do you recommend to run the previous version of the pipeline?

Jose Espinosa-Carrasco

03/27/2025, 8:41 AM

Sorry, the

schema

is just for validation but you this won’t fix the issue. I was referring that on the PR above this should have been fixed, I will check on my site but in the mean time could you share the params-file/config you used to set the paths to the DBs?

Mauricio Morales

03/27/2025, 1:13 PM

I see, maybe there is my problem. Because I download the

DBs

previously, not using the way that the pipeline set up initially Just in case, I used the following command:

Copy code

nextflow run nf-core/proteinfold -r 1.1.1 -resume --input '/users/mmora30/mmora30/workspace/alphafold/alphafold2/alphafold-2.3.1/run_01_h5n1/input/testing_files/samplesheet.csv' --outdir '/users/mmora30/mmora30/workspace/alphafold/alphafold2/alphafold-2.3.1/run_01_h5n1/output/testing_files_out' --mode alphafold2 --alphafold2_db '/users/mmora30/mmora30/workspace/alphafold/full_dbs' --full_dbs true --alphafold2_model_preset multimer --use_gpu true -profile singularity

So, I should say that I did not use a

params-file/config

to set the paths. Maybe that was my mistake? However, I do not think so because of the fact that I pre-downloaded the

DBs

What do you think?

Jose Espinosa-Carrasco

03/27/2025, 3:05 PM

No, this shouldn’t be the issue, I will try to test

1.1.1

but in the meantime you could try to use

dev

version (

-r dev

)

Mauricio Morales

03/27/2025, 3:08 PM

I will try now, thank you for your quick response

Mauricio Morales

03/27/2025, 3:54 PM

It is the same error:

Copy code

E0327 15:43:53.966314 140315935024960 templates.py:837] Could not find CIFs in ./mmcif_files
  Traceback (most recent call last):
    File "/app/alphafold/run_alphafold.py", line 570, in <module>
      app.run(main)
    File "/opt/conda/lib/python3.11/site-packages/absl/app.py", line 312, in run
      _run_main(main, args)
    File "/opt/conda/lib/python3.11/site-packages/absl/app.py", line 258, in _run_main
      sys.exit(main(argv))
               ^^^^^^^^^^
    File "/app/alphafold/run_alphafold.py", line 467, in main
      template_featurizer = templates.HmmsearchHitFeaturizer(
                            ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    File "/app/alphafold/alphafold/data/templates.py", line 838, in __init__
      raise ValueError(f'Could not find CIFs in {self._mmcif_dir}')
  ValueError: Could not find CIFs in ./mmcif_files

Do you think that the fact that the

cif

files are

.gz

play a role in all of this?

Mauricio Morales

03/27/2025, 4:19 PM

One comment from another repository that use

AF

solved related issue editing the path to the

DBs

In that case please edit the path in the variable
template_mmcif_dir
in
run_alphafold.sh

If that is the case, where I can find the files that define the paths for the main pipeline? main.nf? any directory in particular?

Jose Espinosa-Carrasco

03/27/2025, 4:57 PM

Do you think that the fact that the
cif
files are
.gz
play a role in all of this?

What do you mean?

Mauricio Morales

03/27/2025, 5:09 PM

I have the directory like this:

Copy code

├── pdb_mmcif
│   ├── mmcif_files
│   │   ├── 3wae.cif.gz
│   │   ├── 3wt1.cif.gz
...
└── obsolete.dat

Could play a role in the way that the files are detected?

Jose Espinosa-Carrasco

03/27/2025, 5:20 PM

yes definitely, you should have:

Copy code

ls ~/db/alphafold/pdb_mmcif/mmcif_files/ | more
100d.cif
101d.cif
101m.cif
102d.cif
102l.cif
102m.cif
103d.cif
103l.cif
103m.cif
104d.cif
104l.cif

Mauricio Morales

03/27/2025, 5:22 PM

Okay, I will unzip all the files. I will let you know the result!

Jose Espinosa-Carrasco

03/27/2025, 5:23 PM

if you don’t mind could you open an issue in the respository? we may want to support

gz

files, can they directly be swallowed by af2?

Mauricio Morales

03/27/2025, 5:23 PM

No problem, I will do it in a few minutes

thankyou 1

Mauricio Morales

03/27/2025, 6:45 PM

Created! https://github.com/nf-core/proteinfold/issues/279 beer parrot

keanu thanks 1

Open in Slack

Previous Next