Library Prepration for MiniSeq Help

[valenex]

Hello,

I am in a lab that has recently purchased a illumina miniseq. Our main focus for using this machine is to sequence high numbers of PCR products generated from mouse gDNA to genotype specific regions were we have induced a mutation.

Illumina recommended we use the 16S library preparation method and the nextera xt kit for indexing. While this method does work we are facing a pretty significant problem.

Specifically, we generally can receive up to 300 unique samples a week many of which will needed to be amplified with different primers. With the concentration standardization and multiple rounds of purification and PCR before pooling the library. We have found this method to be unreasonably time consuming and wasteful since we will normally need to be working with 100+ samples at a time.

Could anyone perhaps recommend a faster method or one that does not require so many steps as it honestly gets out of control with so many samples with the 16S method.

I can provide more details if I have not been clear enough.

Thank you

[Bukowski]

That library prep is about as simple as it gets. What you need is automation, not a different method.

[thermophile]

I do mostly amplicon sequencing using either one step PCR library generation (targets that we sequence a lot because there's a somewhat substantial investment in indexed primers) following Kozich 2013. For targets that may not justify that expense, I use a 2 step PCR library generation based on Lange 2014. I do have a liquid handler but I think either of those protocols could also be done by hand (using multichannel pipetts) for not much more hands on time.

[Number6]

You can certainly streamline these library preps; we do this for SNP genotyping at dozens or even hundreds of loci in up to 384 samples in a run, using tailed locus-specific primers and the 2-step PCR protocol with dual indexing.

First, you may be able to pool all your locus-specific primers together in the first PCR, especially if you know what you're looking for in the amplicons (e.g. SNP genotyping). It helps to use a master mix made for multiplexing.

Second, you don't need to do any cleanups/quants before going into the second PCR. We simply dilute the first reaction and add a small volume into the indexing reaction. Do a test to find a dilution that works well for you.

Third, if you're doing this repeatedly, you can array the indexing primers in 4 96-well plates (for the 384 combinations) so it's easy to set up the indexing reactions, especially if you have liquid handlers.

Fourth, pool everything and do one cleanup at the end.

As always, it depends on what you're looking for and your tolerance for some dropouts, etc. For us, it doesn't matter if one locus gives us thousands of reads, and another gives us dozens of reads. We can call the SNPs.

[valenex]

Thanks, I've been able to get useful information from all your replies.

[kmcarr]

Fourth, pool everything and do one cleanup at the end.

Lots of excellent advice from Number6. We do a ton of highly multiplexed metagenomic amplicon sequencing in our lab and would add one extra tip for after the final PCR step, before pooling. We use Invitrogen SequalPrep DNA Noramlization plates post PCR to provide some balance to the samples going into the pool. Attempting to quant and balance each PCR rxn individually is an onerous task.

Recovered products from the SequalPrep plates are pooled and as per Number6 a single cleanup is done. QC & quant the pool then onto the MiSeq.

[valenex]

Thanks, any suggestion on what method is most effective for the pooled clean up? I feel a spin column would result in some loss.

[SNPsaurus]

If you are going to have that sort of throughput on a regular basis, you might find it worthwhile to invest in something like the Fluidigm Juno, which lets you load a plate of samples and a plate of primers and does a lot of the work for you.

[thermophile]

I like spri bead cleanup. Theoretically you could use that to normalize as well but I've never had good luck with that and normalize based on either post-clean up picogreen or running on our gel-like detection system. both of those allow for whole plate processing.