Reference path not found in graph

[Karimelakoum]

Hello all,
I would like your help with "Reference path not found in graph" during vg call. While there are previously available answers on the matter, it seems to get me more confused as there isnt a clear answer.
This is a detailed walk though of how I reached the error.
1- apptainer exec pggb.sif pggb
-i genomes/cleaned_genomes.fa.gz
-o fixed_header
-t 8
-p 90
-s 5000
-n 2
-k 19
-V "Dpulex#1#"

WORKS FINE!

2-apptainer exec ../pggb.sif vg autoindex
--workflow giraffe
--prefix daphnia_fixed
--gbz daphnia_fixed_headers.gbz
--threads $SLURM_CPUS_PER_TASK

WORKS FINE!

3-apptainer exec ../pggb.sif vg giraffe
-t $SLURM_CPUS_PER_TASK
-Z daphnia_fixed_headers.gbz
-d daphnia_fixed.dist
-m daphnia_fixed.shortread.withzip.min
-z daphnia_fixed.shortread.zipcodes
-f Dappu_419_R1.fq.gz -f Dappu_419_R2.fq.gz
-o gaf | gzip > Dappu_419_fixed.gaf.gz

WORKS FINE!

4-apptainer exec ../pggb.sif vg pack
-x daphnia_fixed_headers.gbz
-a Dappu_419_fixed.gaf
-o Dappu_419_fixed.pack
-t $SLURM_CPUS_PER_TASK

WORKS FINE!

5-apptainer exec ../pggb.sif vg snarls daphnia_fixed_headers.gbz > daphnia.snarls

WORKS FINE!

6- apptainer exec ../pggb.sif vg call daphnia_fixed_headers.gbz -t 8 -k Dappu_419_fixed.pack -r daphnia.snarls -p Dpulex#1#Chr1#0 -az > Dappu_419_Chr1_final.vcf

ERROR:[vg call]: Reference path "Dpulex#1#Chr1#0" not found in graph

What I do next is check
apptainer exec ../pggb.sif vg paths -M -x daphnia_fixed_headers.gbz

This where I am currently stuck!!. I am confused on why all the chromosomes are marked as HAPLOTYPE when I specifically set Dpulex as my backbone using -V "Dpulex#1#" in command 1.
How can I proceed to successfully do variant calling?
Thanks a lot!
Karim

[jltsiren]

How did you build the GBZ graph daphnia_fixed_headers.gbz? As far as I know, PGGB does not output GBZ directly, and the path metadata conventions it uses in the GFA output are not fully compatible with ours. In particular, because "reference paths" do not exist as a concept in the GFA specification, there are no guarantees that such information can be passed from one tool to another.

If you want to set a sample as a reference in a GBZ graph, you can use:

vg gbwt -Z input.gbz --set-reference sample_name -g output.gbz

And add option --gbz-format if you are using vg version 1.70.0 or earlier.

[Karimelakoum]

Dear,
Thank you for the quick response. Using --set-reference did work!. As the vcf file is now relative to the reference chosen, how do you recommend tracing back to which chromosome the ALT variant is present on. I highly appreciate it!

[jltsiren]

What do you mean by your question? The names of your chosen reference paths are used as contig names in the resulting VCF file. If your path naming scheme uses chromosome names as the sequence/locus name component, you can find the information there.

[Karimelakoum]

As an example in the attachment, the ALT variant in this case is a C instead of a T which should be located on one of the contigs labeled as HAPLOTYPE (alt paths). I was wondering how I can trace back to which contig this ALT variant is carried on.
Note, according to the output of this example T is located on Chr1 but that does not mean the the variant C is also located on the same chromosome correct?