1500 amplicons in salmon

[MattB]

Hi all,

Basically my goal is to 'genotype' 1500 loci in salmon through sequencing. Had initially though of RAD/GBS/ddRAD/etc, however then I realised that one can in theory just do targeted amplicon sequencing with this number of targets. Ultra-high multiplexing with Ampliseq or Truseq custom enrichment seem feasible, however all the assay design tools seem oriented to humans. In theory I could probably multiplex 1000 samples per lane and get 100x coverage of my regions on the Hiseq and also get very efficient sequencing with Proton, but:

1. No targeted enrichment/amplicon kits seem to offer more than 96 barcodes
2. All the assay design tools seem to be for human genome sequence only.

In principle I can't see any reason why Ampliseq couldn't be used in this case if we had the sequence available...or am I missing something here?

Has anyone else had experience with sequencing of 1000+ amplicons in animals/plants with high level multiplexing/barcoding?

Cheers,
Matt

[rosatoc]

Hi Matt
How is your GBS in salmon project going? Have you found or developed tools?
Thanks

[MattB]

Still chewing over various options...in principle it should be possible with 'Ampliseq' or Fluidigm Access Array too (documented 10 amplicon multiplex x 48 so 480 potential amplicons per chip). Ampliseq officially only supports human and mouse so we have to see how we could customise this to fish.. Basically we just need to be able to run an ultra-high PCR multiplex, which is really just a matter of primer design and PCR chemistry I guess

To get the costs down, we need to be going for multiplexing levels of at least 384 per lane on Hiseq, for example, but with only this small number of amplicons we would have plenty of coverage and could mutliplex even higher. Seems that people are multiplexing to that level now more commonly with custom barcode combinations etc. so that shouldn't be too big a challenge.

So still tossing up a few options...RAD/GBS/ddRAD is still an option too of course but I am keen to target my selected amplicons rather than a random sample, and also I might have some quite degraded samples that will be problematic with any RAD type approach, but shouldn't be an issue for PCR of small amplicons.

Of course we have array genotyping options with Illumina and Affymetrix, but the sequencing option with a high level of sample multiplexing it could be extremely efficient (both time and cost).

[Vinz]

Multiplexing should be no problem for Illumina with dual indexing. We have 1536 barcodes in use without problems: 96 x 16. We used so far only 480 in a single run, but that should not make a difference...

[MattB]

Thanks Vinz, so the mutliplexing/barcoding isn't an issue, just the super-high PCR multiplexing challenge remains (if we want 1000+ amplicons). However, with a high-fidelity enzyme like the NEB Q5 or Phusion and careful primer design...wonder how far we could go with just trying such a PCR and seeing what happens Can't be anything else that magical about Ampliseq, it seems to just be a multiplex PCR with thousands of primers...

[JackieBadger]

We sequence over 1000 individuals per run using duel MIDs.
You could increase this further by allocating different Illumina indices to different pools that use the same MID combinations. So without depth being a limiting factor you could easily sequence 1000s of samples per run.

[rosatoc]

Hi Vinz and Jackie
Are you talking about bead chips or about high throughput sequencing? I have a goal to begin offering Genotyping by Sequencing to the researchers this genomics core supports, which means I am not focused on a specific organism but focused on providing a high value technique, on the HTS platform.
Hi Matt
Are your markers microsattelites, SNPs, or ?
Thanks : )
Caprice

[JackieBadger]

I am talking Illumina Miseq amplicon sequencing

[MattB]

Caprice, I am talking about SNPs. We could of course 'genotype' with Affymetrix/Illumina/Sequenom/Fluidigm etc., or do GBS/RAD/ddRAD sequencing as a random sampling of SNPs, but I am particularly interested in NGS of a set of 1500 or so amplicons (SNPs) over several thousand samples.

I am talking Illumina Miseq amplicon sequencing

JackieBadger, how many amplicons are you sequencing, just a handful or a large set?

Matt

[JackieBadger]

Lots of alleles within individual amplicons, but we amplify these using just one degenerate primer set...so its quite a simple set up.

[MattB]

Thanks, wondering how far we could push a PCR multiplex if we tried to include hundreds of amplicons...

[adrien.andre]

Hey Vinz, could you please give me more information about this?
I'd really like to know the index sequences
here is my address :
adrien.andre
at
ulg.ac.be

thanks!

[Vinz]

We published the indexing for 1536 samples in the Supplementary of
this article in BMC Genomics.

[apredeus]

We published the indexing for 1536 samples in the Supplementary of
this article in BMC Genomics.

Wow, that's a pretty cool method you have there. Congrats on a nice paper.