Genome sequencing project opinions & help - heterozygous species

Question

[lessismore]

Hi all,

as here there are notably many experts in this field i am curious to have your opinion about a delicate and arguable topic.

According to the current technologies, what kind of sequencing (technology/platform, read length and coverage) do you suggest for efficiently sequencing and, hopefully assembly, the genome of a plant species of about 1,2 Gb and highly heterozygous ?

Thanks for your help

Answer 1

[shwethacm]

Hi there,

To begin with, highly heterozygous genomes are not the easiest to assemble using genome sequencing technologies and assemblers that we have today. Be prepared to get a rather fragmented assembly with low-moderate N50.

I would suggest a combination of short-read (Illumina) and long-read (PacBio) technologies for your genome. Most of the recent large-genome assemblies have adopted this combination and it has worked well in almost all cases. Build the skeleton genome assembly using long reads, and use the short read data to "fill-in" and polish your assembly.

If you only want to use short-reads, Redundans is an assembler that works with short-read data and is specialized for highly heterozygous genomes. You could plan your assembly project using short reads alone with this assembler in mind. Plan your sequencing depth and insert sizes according to the recommended values so you have the right 'type' of data before you assemble.

Also, this might be of interest to you: This preprint that just came out describes the sequencing ans assembly of the wheat genome, which is highly heterozygous and massive (>10GB genome size). http://www.biorxiv.org/content/early/2017/07/03/159111

Good luck!