Unable to apply custom grouping in histgrowth

[KopalliV]

Hi,

I am running Panacus on a merged chromosome-wise pangenome graph. The graph has 220 paths: 22 genomes × 10 chromosomes. I want to group chromosome paths by genome, so growth/core analysis should use 22 groups, not 220 paths.

My custom grouping file looks like this:

Chr01_Sb_HMW_1#0#Chr01_Sb_HMW_1#0 Sb_HMW_1
Chr02_Sb_HMW_1#0#Chr02_Sb_HMW_1#0 Sb_HMW_1
Chr03_Sb_HMW_1#0#Chr03_Sb_HMW_1#0 Sb_HMW_1
...
Chr01_Sbi_v5_ref#0#Chr01_Sbi_v5_ref Sbi_v5_ref
Chr02_Sbi_v5_ref#0#Chr02_Sbi_v5.0_ref Sbi_v5_ref
...

The file has 220 rows and 22 unique groups.

I tried: panacus histgrowth -t 32 -c bp -l 1,1 -q 0,1 -g panacus_groups.fixed.tsv ../SBi_21_mc.og.gfa > SBi_21_mc.grouped_growth.bp.tsv

Expected: growth curve should go from 0 to 22 after grouping.

Observed: growth curve still goes from 0 to 220, suggesting grouping is not being applied to the growth units.

Could you please clarify whether the custom grouping file format is correct and help me understand?

Panacus version: 0.5.0

Thanks

[danydoerr]

Hi @KopalliV, make sure your file is tab-separated. Other than that, it looks fine.

[KopalliV]

Thanks. I confirmed that the grouping file is tab-separated, has 220 rows, and 22 unique groups.

Example:

Chr01_Sb_HMW_1#0#Chr01_Sb_HMW_1#0 Sb_HMW_1
Chr02_Sb_HMW_1#0#Chr02_Sb_HMW_1#0 Sb_HMW_1
Chr03_Sb_HMW_1#0#Chr03_Sb_HMW_1#0 Sb_HMW_1

However, after rerunning:
panacus histgrowth -t 32 -c bp -l 1,1 -q 0,1 -g panacus_groups.fixed.tsv ../SBi_21_mc.og.gfa > SBi_21_mc.grouped_growth.bp.tsv

the output still goes from 0 to 220, not 0 to 22.

So the file appears to be tab-separated and formatted correctly, but histgrowth still seems to use the original path count as the growth unit.

[heringerp]

Hi @KopalliV

thanks for the notice. There was an issue in the grouping code, that accidentally did group by haplotype instead of using the provided file. I fixed the issue now on the main branch. I will push a new release for this on Monday, but it will probably take a couple more days till the update is on bioconda.

[heringerp]

Fix is now also available with conda as v0.5.2