Failure of `panaroo-merge` with `-a core`

[HarryHung]

I was trying to run panaroo-merge with -a core to generate output for Pyseer.

panaroo-merge -d [lots of directories here] -o results -t 32 -a core

However, it failed at the final stage with the following error:

Traceback (most recent call last):
  File "/usr/local/bin/panaroo-merge", line 10, in <module>
    sys.exit(main())
  File "/usr/local/lib/python3.10/site-packages/panaroo/merge_graphs.py", line 566, in main
    merge_graphs(directories=args.directories,
  File "/usr/local/lib/python3.10/site-packages/panaroo/merge_graphs.py", line 422, in merge_graphs
    generate_core_genome_alignment(G, temp_dir, output_dir, n_cpu, alr,
TypeError: generate_core_genome_alignment() missing 1 required positional argument: 'hc_threshold'

If I am not mistaken, it seems to be relevant to the following lines which only passing 9 arguments to the function:

panaroo/panaroo/merge_graphs.py

Lines 422 to 424 in 18d7c9d

	generate_core_genome_alignment(G, temp_dir, output_dir, n_cpu, alr,
	isolate_names, core, len(isolate_names),
	hc_threshold)

While the function is expecting 10:

panaroo/panaroo/generate_output.py

Lines 433 to 435 in 18d7c9d

	def generate_core_genome_alignment(
	G, temp_dir, output_dir, threads, aligner, isolates, threshold, codons, num_isolates, hc_threshold
	):

As the arguments are positional instead of keyword, the actual problematic arguments should be core in merge_graphs.py vs threshold + codons in generate_output.py, instead of hc_threshold is missing.

Thanks, and I hope there will be a fix for it soon!

[nzmacalasdair]

Hi Harry,

Thanks for flagging this. It looks like the alignment code from panaroo-merge hasn't been updated to include the production of a filtered core alignment, which is causing this error.

This should be a quick fix, but depending on how you've installed panaroo it might take a while to reach you. Given how long the merge is likely to take, the fastest thing to do is definitely to run the merge without the alignment as a first step, ie: panaroo-merge -d [lots of directories here] -o results -t 32 , and then run the core alignment separately on those results, with panaroo-msa -o results -t 32 -a core. This has the added advantage of splitting it across two jobs, if you are running panaroo on a cluster with short runtime limits.

Let me know if any of this is unclear, or if there's anything else I can do to help!

[HarryHung]

Oh cool, did not know I could run the alignment separately.
I have already done the merge only before, was re-doing the whole thing just for the core alignment.
I will give it a try. Thanks!

[nzmacalasdair]

Fixed with commit 7ef8576