Failed to find source directory "/home/gpux1/alphafold/uniref90" to mount in Docker container.

[Lili-irtyd]

Hi, everyone. I have followed all the steps as instructions, but when I execute the last step(I have just installed the reduced_dbs)

python3 docker/run_docker.py \
>   --fasta_paths=T1050.fasta \
>   --max_template_date=2020-05-14 \
>   --model_preset=monomer \
>   --db_preset=reduced_dbs \
>   --data_dir=$DOWNLOAD_DIR \
>   --output_dir=/home/user/absolute_path_to_the_output_dir

The traceback is as follows

 File "/home/gpux1/alphafold/docker/run_docker.py", line 267, in <module>
    app.run(main)
  File "/home/gpux1/miniconda3/lib/python3.11/site-packages/absl/app.py", line 312, in run
    _run_main(main, args)
  File "/home/gpux1/miniconda3/lib/python3.11/site-packages/absl/app.py", line 258, in _run_main
    sys.exit(main(argv))
             ^^^^^^^^^^
  File "/home/gpux1/alphafold/docker/run_docker.py", line 210, in main
    mount, target_path = _create_mount(name, path)
                         ^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/gpux1/alphafold/docker/run_docker.py", line 114, in _create_mount
    raise ValueError(f'Failed to find source directory "{source_path}" to '
ValueError: Failed to find source directory "/home/gpux1/alphafold/uniref90" to mount in Docker container.

Not sure how to fix this problem, is it need to install the uniref90 first?

Thanks!

[lan99652]

I have also encountered this situation. Have you resolved it?

[Lili-irtyd]

I have also encountered this situation. Have you resolved it?

I don't resolve it, but I think maybe the reason is lack of enough hard disk space. If you resolve it, please tell me your solution, thanks!

[tcoates5]

When you say you followed all the steps, did you run ./download_all_databases.sh $DOWNLOAD_DIR reduced_dbs, and did it finish without raising an error? Assuming you are running in a linux environment, you can try tools like df to see how much space you have available in case you can't easily check to see if the download script errored out

EDIT TO ADD: I have experienced running out of space in the middle of a run, but it never raised an error that looked like the one you shared. Also, as far as I understand it, the main difference with reduced_dbs is that it uses the smaller bfd database, pretty sure it still uses the other databases along with it. If you do have the appropriate databases downloaded, the next thing to check is permissions on those files, as error messages that say "Can't find" are often really "Can't access."

[Lili-irtyd]

When you say you followed all the steps, did you run ./download_all_databases.sh $DOWNLOAD_DIR reduced_dbs, and did it finish without raising an error? Assuming you are running in a linux environment, you can try tools like df to see how much space you have available in case you can't easily check to see if the download script errored out

EDIT TO ADD: I have experienced running out of space in the middle of a run, but it never raised an error that looked like the one you shared. Also, as far as I understand it, the main difference with reduced_dbs is that it uses the smaller bfd database, pretty sure it still uses the other databases along with it. If you do have the appropriate databases downloaded, the next thing to check is permissions on those files, as error messages that say "Can't find" are often really "Can't access."

I did run ./download_all_databases.sh $DOWNLOAD_DIR reduced_dbs, it firstly rasied that (OK)the download is completed, but then it raises the problem is "alphafold/DOWNLOAD DIR/small bfd ~/alphafold gzip: DOWNLOAD_DIR/small bfd/bfdfirst_non_consensus_seguences.fasta.gz: No such file or directory". Maybe the reason is this, but I don't know why it could happen, because it seems that I have this file bfdfirst_non_consensus_seguences.fasta.gz.

Also, I want to ask it when you run ./download_all_databases.sh $DOWNLOAD_DIR reduced_dbs, is it appear to have 2 documents named bfdfirst_non_consensus_seguences.fasta.gz and params.gz( I don't remenber the suffix), is it correct?

Thanks!

[tcoates5]

Since you have the file that the download script said didn't exist when it went to decompress, you probably need to change the permissions on the tarball (probably using chmod) then manually run gzip. Based on what you have said so far, odds are good that, as a security measure on the machine you are putting this on, downloaded files automatically have restricted permissions. As a result, you will probably have to go through the process of using commands from each of the scripts that are normally called for you by download_all_databases.sh, as you should definitely have more files than that. To save yourself time, I recommend using something like the screen command or multiple terminals to set up the downloads in parallel.

[sherryliu987]

If you're struggling to install AlphaFold locally, feel free to try https://www.tamarind.bio/alphafold, a website which offers a no-code interface for bioinformatics tools including protein structure prediction with AlphaFold for free.