Amplicon Sequencing & Reads Per Sample (Coverage)

[cement_head]

Hello,

Is there a "best" way to normalise and pool amplicon samples to acquired even, or balanced read coverage. Otherwise known as getting the same "reads per sample" for each of the pooled samples? We often end up with stochastic coverage despite our best efforts. I'm looking for a better way...

Thanks,
Andor

[GenoMax]

Can you add information about the size ranges of amplicons?

[cement_head]

Hello,

For example, 16S V4: about 350 bp (total length). We are using the Earth Microbiome Project protocol.

Regards,
Andor

[kmcarr]

Originally Posted by cement_head

Hello,

Is there a "best" way to normalise and pool amplicon samples to acquired even, or balanced read coverage. Otherwise known as getting the same "reads per sample" for each of the pooled samples? We often end up with stochastic coverage despite our best efforts. I'm looking for a better way...

Thanks,
Andor

Don't focus as much on "same reads per sample" but more on "enough reads per sample" even if the sample-to-sample count varies.

When dealing with 100's of samples in a typical metagenomic experiment, individually quantifying and normalizing PCR products is unrealistic. Our lab opts for using Invitrogen SequalPrep DNA Normalization plates (https://www.thermofisher.com/order/c...oduct/A1051001). After running your PCR you dilute it in the SequalPrep binding buffer and add it to the plate. The wells are coated with something which binds dsDNA; the binding capacity is limited and your PCR products are theoretically in excess. After a short incubation you remove the excess DNA, wash and elute the bound DNA. Pooling is done as the eluted samples are collected. We do a final cleanup on the pool, QC, quantitate and sequence.

Real life experience -- the amount of DNA recovered per well is not what their spec sheet says it will be and we still may see variation of ~4-5X in the number of reads per sample (though not always this much). Also, these are pricey, about 110 USD per plate (a little over $1.00/sample). Even with these drawbacks it is a very fast and easy method to (somewhat) normalize your metagenomic libraries.

[thermophile]

my facility has a capillary fragment analyzer (qiaxcel, ~$0.40/sample not considering machine cost). I pool eqimolar based on that, then clean the entire pool in a single spri cleanup. I'm generally happy if I don't have many samples with an order of magnitude more sequences than I'm aiming for. I didn't buy the qiaxcel and if I didn't have it, I'd probably do what kmcarr does. I totally agree that I'm concerned with getting enough reads per sample and don't worry about the few winners of the pooling lottery. I run on MiSeq so the few samples that don't get enough seqs can just be repooled for the next run.

[acockburn]

Sequal prep
https://www.thermofisher.com/order/c...oduct/A1051001

This is how we normalize.

[cement_head]

Originally Posted by acockburn

Sequal prep
https://www.thermofisher.com/order/c...oduct/A1051001

This is how we normalize.

Yeah, I'm going to check these out...