Hi, I sequenced some bumblebee genomes (short-read), followed by pooling the samples and generating a genome assembly using ABySS. It is obviously a fragmented assembly with short-read data but when I used GenomeScope2.0 to obtain a genome size, it gave me 120 mb, which is around half of what I got from ABySS (235 mb).The organism is diploid. What could be a reason for this discrepancy?
Cheers, Saam
My bet would be that the additional sequence is haplotypic variation that couldn't be collapsed. Are the sequencing data coming from a single individual bumble bee, or have you extracted genomic DNA from multiple ones? Because if you used multiple individuals from a non-inbred population, then heterozygosity would probably be high and produce an extra-fragmented assembly (i.e. more fragmentation than you would expect because you're using short reads).
was thinking along the same lines (and is indeed the most likely first 'guess'). On the other hand: genomescope gives some indications about haploid/diploid size, no?
@ saamhasan55 : can you post some more numbers/plots from the genomescope analysis?
good point.
and while we're at it, I would consider downsampling (to 50-70x) the input data (eg. run BBnorm on them), 265x is quite a huge coverage, it might start confusing rather than helping the assembler at some point.
That seems a good idea, thanks. I will try that.
Log in to answer this question.
do you have any prior knowledge on the expected genome size?