Error in ppanggolin msa when running all partitions

[ktmeaton]

I'm getting an error when I try to run ppanggolin msa with --partition all:

2024-03-20 17:46:53 writeMSA.py:l214 INFO       Computing the MSA ...
Traceback (most recent call last):
  File "/home/username/.conda/envs/ppanggolin-2.0.4/bin/ppanggolin", line 10, in <module>
    sys.exit(main())
  File "/home/username/.conda/envs/ppanggolin-2.0.4/lib/python3.10/site-packages/ppanggolin/main.py", line 197, in main
    ppanggolin.formats.writeMSA.launch(args)
  File "/home/username/.conda/envs/ppanggolin-2.0.4/lib/python3.10/site-packages/ppanggolin/formats/writeMSA.py", line 348, in launch
    write_msa_files(pangenome, args.output, cpu=args.cpu, partition=args.partition, tmpdir=args.tmpdir,
  File "/home/username/.conda/envs/ppanggolin-2.0.4/lib/python3.10/site-packages/ppanggolin/formats/writeMSA.py", line 325, in write_msa_files
    compute_msa(families, outdir, cpu=cpu, tmpdir=tmpdir, source=source, use_gene_id=use_gene_id, code=code,
  File "/home/username/.conda/envs/ppanggolin-2.0.4/lib/python3.10/site-packages/ppanggolin/formats/writeMSA.py", line 215, in compute_msa
    bar = tqdm(range(len(families)), unit="family", disable=disable_bar)
TypeError: object of type 'generator' has no len()

I think the issue might be that when you use the all partition, the gene families are a generator in get_families_to_write. But for any non-all partition (ex. core), the gene families are a set:

PPanGGOLiN/ppanggolin/formats/writeMSA.py

Lines 57 to 61 in 9e9acdf

	families = set()
	nb_org = pangenome.number_of_organisms
	if partition_filter == "all":
	return pangenome.gene_families

Which then fails when it tries to call the len method on a generator in compute_msa:

PPanGGOLiN/ppanggolin/formats/writeMSA.py

Line 213 in 9e9acdf

bar = tqdm(range(len(families)), unit="family", disable=disable_bar)

If I change the return value to a set in get_families_to_write, it seems to work!

if partition_filter == "all":
    return set(pangenome.gene_families)

Is this a reproducible error for other users, and does this solution seem feasible? Thanks!

[axbazin]

Hi,

The problem is indeed reproducible, and the solution is exactly the one we're going for.

It will be coming up as soon as we merge #196 and get it to the next patch release.

Thank you for the bug report !

Adelme

[ktmeaton]

Thank you!

[JeanMainguy]

The fix has been released in version 2.0.5.
It should be available in bioconda very soon.