kmer analysis for genome size estimation

Question

[rj.rezwan]

Hi, I used the PacBio HiFi reads in Merqury/1.9 for kmer analysis of our allotetraploid species. I have shared the figure of genomescope plot, which showing the size of 1.5gb. I am astonished what's wrong here. Can someone guide me in this? For more information, the diploid species has a genome size 1.5 gb. I shall be grateful to you. First, I assembled the genome using hifiasm with size of ~3.7gb.

Second question is why the observed peaks are going out of the model peak?

[Michael]

Isn't this size estimate a good fit? I am not sure, but is Ploidy taken into account at all for this? You can also see that dup:4.11 which would fit with tetraploidy. Also, even if there was very high heterozygosity in the genome, this would at best add a fraction of a percent more unique k-mers.

[rj.rezwan]

yes, this thing want to ask why this does not give a length near to my assembly size? and please look at the observed peak (blue), is this normal to be out of the model peak (black)?

Answer 1

[Philipp Bayer]

The oh-know materials have all kinds of example plots for various genome cases or extra settings for cases where the size estimation didn't work out. In your case, looks a little bit like the allotetraploid example plot? https://github.com/KamilSJaron/oh-know/wiki/Allotetraploid

[rj.rezwan]

yes, genome is allotetraploid