Entering edit mode
4.9 years ago
dongseokkim6662
•
0
When I tried to use default option of jellfyfish and GenomeScope, it could not make the result. It said that profile cannot converge... So, I tried to adjust the kmer value and Max kmer coverage, such as 17 kmer, Max kmer coverage = 1,000,000. And then the results was emerged. But, I think this results is weird. Is it possible to make negative value of unique length?
You can check my results in below link.
http://genomescope.org/analysis.php?code=0b75l7MdAaX4mz64TSKa
yes, you can play (tweak) the params a little, and sometimes you will indeed have to even. However I'm afraid the parameters you mention here are not making much sense. Often you might also simply not have enough data to do the maths on (eg. coverage too low).
anyway, from the output you get you could have noticed your parameters are not making much sense, as you get a negative! fraction of uniq sequences in your dataset apparently ;)
Thanks for replying my question. :)
Can you tell us what kind of sequencing this is? What is the expected genome size? How much sequence you have?
I did Illumina Novaseq6000. The expected genome size is about 20~30Mb because the genome size of the other species which is same genus is about 25Mb. And I sequenced the species about 26Gb.
so you sequenced it ~1040x coverage ?
yes, we usually sequenced samples about 20Gb (illumina)
that's a serious overkill for such small genomes and will likely rather hinder the assembly process than improve it. Downsampling might be an advisable strategy here.
What you do you mean by 'usually' ? I would assume you estimate your seq-depth by the genome specs you want to sequence, no?