Kmer genome size estimation so much lower than the actual

[GAFA]

Hi everyone,

I’m assembling an eukaryotic genome (2n=22) for the first time, working in a non-model plant species, and I could use some insight: my data consists of reads from a full lane of Illumina HiSeq 2x151 sequences with insert size ~350. from several literature, the estimation of haploid genome size of this plant was about 400Mb. However, kmer-counting programs such as Jellyfish have predicted an assembly size of less than half that number, at about 210Mb.

Does anyone have any idea why the nuclear genome size is so much larger than that of kmer one?

Any related thoughts/comments would be, by me, appreciated!

[Brian Bushnell]

It would be nice if you could post the kmer frequency histogram (both as text and as an image). Kmer-counters can make mistakes on genome size estimation when the data is noisy, or if the peaks are too broad, or if they guess the ploidy wrong, etc. I'd be interested in seeing what BBMap's KmerCountExact reports as the estimated genome size, also:

kmercountexact.sh in=reads.fq khist=khist.txt peaks=peaks.txt

[GAFA]

Dear Brian,
I have estimated kmer count using Jellyfish.
hereunder, the links to histo image and text.

https://www.dropbox.com/s/ar60gcuiqsn7fbv/histo.pdf?dl=0

https://www.dropbox.com/s/wse5iu02biv93ad/histo.txt?dl=0

I will try using BBMap and let u know

thanks

[Brian Bushnell]

That looks like a haploid to me. There is no evidence from the kmer frequency histogram that the organism is a diploid. There is exactly one prominent peak, and it is very clear. Is this sample highly inbred, or wild-type?