I am stuck with a de novo genome assembly. I have done a lot of QC filtering :
1) quality and adapter trimming
2) removing contamination
3) removing duplication
I always end up with that kind of k-mer spectrum (and a bad N50):
based on that paper:
I know that I should end up with a mono or multimodal distribtution.. which is not the case.
If you guy have some idea, would be helpful.