Combine TruSeq DNA PCR-Free library and 2x300bp PE sequencing on MiSeq

[Johnny99]

Hi everyone,

We are about to do whole genome sequencing of an 12Mb organism. There are still a lot of "loose ends" in the reference genome, so we aim for long read lengths in order to contribute to the assembly.

Does anyone have experience in 2 x 300bp PE sequncing on the MiSeq, using the TruSeq DNA PCR-Free Library Prep kit (550bp fragments)? Our concern is that the shortest fragments hybridize more easily to the flowchip causing a lot of redundant reads. There will of course be an overlap of 50bp for the 550bp fragments, but will the problem of big overlaps be extensive?

Has anyone used alternative protocols with a more strict size selection or adjusted settings for the fragmentation to get less short reads?

Any help is greatly appreciated!

[nucacidhunter]

If you use standard bead-based size selection approximately half of reads will have length below 550 bp (see PCR-free user guide plots). Two options:

1- Use more input DNA for PCR-free and size-select with Pippin and follow the rest of protocol
2- PacBio, with 12Mb genome you can get 100x coverage with 2-3 SMRT cells which would give better results than any Illumina system for less cost

[kerplunk412]

2- PacBio, with 12Mb genome you can get 100x coverage with 2-3 SMRT cells which would give better results than any Illumina system for less cost

I completely agree with this. If you are looking to "finish" a genome I believe PacBio is generally accepted to be the gold standard.

[Brian Bushnell]

I completely agree with this. If you are looking to "finish" a genome I believe PacBio is generally accepted to be the gold standard.

That's absolutely true. Pure PacBio is currently the cheapest and most effective way to produce a high-quality single bacterial genome. That said, I've never heard of a prokaryote with a 12Mbp genome... so I'd really need to know what this organism is before giving a firm recommendation.

Aside from PacBio, you can still get an extremely good assembly for most bacteria using Illumina 2x300 reads. Depending on the assembler, it may be useful to merge reads before assembling. You will not get a 1-contig assembly from an Illumina fragment-only library on a bacteria, regardless of the insert size, but the result will typically still be very good for most purposes.

[Johnny99]

Thanks a lot for your considerations!

The organism is an eukaryotic parasite. Our primary goal for the experiment is variant calling, so we aim for good coverage. However, it would be a bonus if we could contribute to improve the reference genome at the same time. That is why we consider long reads. That might be too much to ask for...