issue with assembly of plant genome sample

[dasn588]

Hi,
I am facing issue in assembling a plant genome with expected genome size of around 1 GB. I have around 6X: Duplex data, 12X: UltraLong data and 35X: Pore-C data. I am getting assembly of only 60MB around as final output.
I am using following commands for Verkko assembly;
./verkko -d Assembly_Verkko/ --hifi duplex_porechop.fastq --nano ultralong_porechop.fastq.gz --porec porechop.fastq.gz --telomere-motif TTTAGGG --local-memory 800 --local-cpus 80

I am sharing the snakemake log file for given process.

Please let me know the following things:
if the tool is installed properly
if i am running the program correctly
Also, how to resolve the issue to get correct assembled genome fasta

2025-01-17T073408.280509_snakemake.log

Thanks
Nihar

[skoren]

Is all your high accuracy input coverage 6x? That is definitely too low, assuming you have a diploid that is a per haplotype coverage of 3x which is barely enough to cover the genome once. The 12x of UL is also much lower than we recommend, is that all data or cut at only those > 100kb?

We typically recommend 20-25x/haplotype of high accuracy data (so duplex, hifi, herro corrected) which would be 40-50x for a diploid. For UL we recommend 10-15x per haplotype so 20-30x for a diploid. You could boost your high accuracy coverage by running herro on the UL data over 10kb (assuming it's a recent R10 run) but that would still be barely enough coverage for an assembly.

[dasn588]

Thank you for your response.

But still I did not get response for other two queries;
if the tool is installed properly (based on attached snakemake log file)
if i am running the program correctly

[skoren]

Yes, the command and log look reasonable. You shouldn't need the local resource spec (it will autodetect what is available on the node or what you've allocated via grid request) unless you want to restrict it to a fraction of the machine manually.

[skoren]

Idle