ajasja / af2slurm Goto Github PK

Dear developer,

would it be possible to prepend some kind of identifier to the SLURM job names to easily identify AF2 jobs with the 'squeue' command. Just an idea.

Keep up the good work and thank you!

Sometimes things go wrong, so it would be useful to copy over the fasta file in the in folder into the out folder and append an .original

For example in/test.fasta --> out/test/test/test.fasta.original

Or something similar.

On windows systems the extension is often hidden and files look like they have a .fasta extension but are in fact fast.txt.

It would be great if these files (and perhaps a .txt) would be handled transparently.

If there is an empty line after the comment section the parsing crashes.

For example a fasta file with:

# --msa-mode single_sequence 

>P1__P2
AEIAALEAENAALEAENAALEAEIAALE:
AKIAALKAKNAALKAKNAALKAKIAALK

Crashes with this error message:

 File "af2slurm-watcher.py", line 156, in <module>
    main()
  File "af2slurm-watcher.py", line 148, in main
    move_and_submit_fasta(fasta, args, dry_run=args.dry_run)
  File "af2slurm-watcher.py", line 82, in move_and_submit_fasta
    fasta_path, out_folder=args.out_folder, dry_run=args.dry_run
  File "af2slurm-watcher.py", line 51, in move_over_fasta_file
    if lines[0][0] != ">" and file_path.suffix != ".a3m":

Can we use the exact same code to run other programs, such as codon optimisation. I think it should be possible with just modifying the settings:)