My smugeplot looks strange about smudgeplot HOT 1 CLOSED

Inference of ploidy from sequencing data can be tricky, especially when one relies on assembly.

This species is very strongly supported to be triploid. We see very clear separation of smudges thanks to very deep coverage. This is really good and allows us to be sure that the individual smudges are from different distributions (i.e. produced by different genomic kmer compositions). One should doubt inference when the converge is low, but this is very convincing evidence.

We search for kmers that are 1 SNP away from each other. Therefore there are heterozygous positions that we will miss and also sometimes will happen that two haplotypes have a variant and the third haplotype has more than one. The two will form a pair and the third one won't be recognized due to the substantial divergence. Given this we can guess that two of the haplotypes are fairly same (which is the reason why you can not picked it up in your assembly) and the third haplotype is substantially diverged which let in your assembly to frequent separatedly assembled haplotypes, but always just in two copies.

Does it make sense?

from smudgeplot.