Hybrid strategy? Got IonProton - which MiSeq?

[spabinger]

Hi,

we want to assemble a eukaryote with an estimated genome size of about 22MB.
Currently, we have 108X SE IonProton reads and have the possibility to do some more sequencing on a MiSeq.

Should we go for:
- PE protocol (Nextera - 2x150bp - 250bp insert size) or
- MP protocol (Nextera MP - 3,5,8kb insert size)

What insert size should we choose if we go for MP?

Other options (HiSeq, PacBio, ...) are not available.


Thanks,
Stephan

[Brian Bushnell]

If the current reads are long and good quality, then you probably wouldn't benefit much from an Illumina fragment (PE) library. Go for a MP library with the longest possible insert size to get the best scaffolding and repeat resolution.

[spabinger]

Hi Brian,

thanks for your answer. I also thought that MP would be the best way to go.

Best,
Stephan

[Buzz0r]

what is your average read length on the Proton. I just saw some microbial data sets from a collaborator which clearly showed extreme improvement for the assemblies for the 2x300Bp reads on Miseq (avg 296Bp) as compared to a 200Bp run on Proton with the same samples (avg 163Bp).

MP can help if you have enough reads with somewhat decent quality but always remember that q20 is not q30!

[Brian Bushnell]

MP can help if you have enough reads with somewhat decent quality but always remember that q20 is not q30!

True, but q30 is often not q30, either Most assemblers are tolerant of some level of error, and many (like AllPaths LG) do their own error-correction. There are also standalone error-correction tools such as BBNorm that you can use. And MiSeq 300bp reads will not be >q30 all the way across; the tail will be much lower.

It partly depends on how repetitive the genome is and how large the repeats are. The more repeats, and the longer the repeats, the more valuable LMP is.