Bootstrap/jacknife perl script does not work on large datasets

[mdiricks]

Describe the bug
Bootstrap/jackknife perl script does not work on large datasets

To Reproduce
Run the bootstrap or jackknife perl script on a set of 1800 strains. (scripts work on set of 42 strains)

Expected behavior
A clear and concise description of what you expected to happen.

Screenshots
Error message:
Can't exec "/bin/sh": Argument list too long at /molmyc/miniconda3/envs/mashtree/bin/mashtree_bootstrap.pl line 103.
mashtree_bootstrap.pl: main::main: Died
Stopped at /molmyc/miniconda3/envs/mashtree/bin/mashtree_bootstrap.pl line 104.

Desktop (please complete the following information):

• OS: [e.g. iOS]: Windows (Ubuntu 20.04 LTS
• Version 1.2.0 [e.g. v0.57, can be obtained by running mashtree --version.]
• which method did you install with? Conda

Additional context
Add any other context about the problem here.

[Subhajeet1997]

I am also facing this problem. I am trying to run on 3522 genomes and it throws the above error. Any solution to this?

[lskatz]

Hi, sorry I didn't see this issue until today. I think you are running into the limit of parameters you can use in shell. To overcome this, you will need to create a file of files and then use --file-of-files. For example:

find input -name '*.fasta' > fastas.fofn
mashtree_bootstrap.pl --reps 100 -- --file-of-files fastas.fofn

I didn't personally test this example today, and so please come back and let me know if this worked for you.

[mdiricks]

Hi,
Unfortunately, I get an error: ERROR: Did not find fasta records in "input files" using your solution. the fastas.fofn file contains all the file names.
Could you perhaps already test it?

[lskatz]

Hi, I made a new tag v1.3.0 that hopefully fixes this bug. Please let me know if this helps.