kishwarshafin/helen

Plant species

Closed this issue · 3 comments

Jolvii85 commented

Hi kishwar
Can I use this polisher on complex plant genome?
Does the model trained with human sequencing data work for the plant species?

Thank you.

Jolvii

kishwarshafin commented

Hi Jolvii,

The model will treat the genome as a homozygous sequence. We have not benchmarked any plant genome, but if you have sequencing data and a high-quality assembly that we can use as a ground truth then we can do a quick test and let you know. (None of us are plant genome experts :( sorry)

Best,
Kishwar

reachsagaya commented

Hi Kishwar,

I have a similar question and I work with plant genome as well.
While choosing params.json when running Margin polish, can I use allParams.np.human.guppy-ff-235.json for the plant genome or should I use allParams.np.json.
Next if I shared with you my draft plant shasta assembly using nanopore reads would you be interested in using it to benchmark?
best regards
Sagaya

kishwarshafin commented

@reachsagaya , the choice of the model only depends on what version of guppy basecaller you have used for your data. If you have used Guppy 2.3.5 then you should use np.json.235 where 235 reflects the guppy version number.

If you can provide us a good reference genome and your assembly then I can do a quick test of sequence quality for you. The reference genome is expected to be the same species. Like the same way we use GRCh38 for the human genome.