Balance many PCR products for pooling

[Marius]

Hi everyone!

I'm currently doing a project in which I PCR amplify (long-range PCR) 20 genomic fragments (each of 5-10 kb in length) in 50 individuals. The aim is then to pool fragments of the same individual and mark each such 'pool' with a unique barcode. After doing so, the 50 individually barcoded pools will then be together sequenced on a single Illumina lane.
I'm aware that the best way to quantify PCR products for subsequent balanced pooling would be based on a Bioanalyzer measurement. Nevertheless, this would be too expensive and time consuming given that there would need to be done about 1000 measurements (50 x 20). Another way would be to gel-elute the PCR products and then measure the concentration (e.g. wih a Qubit). But again, this would lead to a total of about 1000 gel elutions...
As the PCRs are not working with the same fidelity I'm looking for a feasible way on how to quantify PCR products to pool them at more or less equal amounts for subsequent barcoding. The balancing doesn't need to be perfect, as I expect a very high coverage due to massive sequencing with Illumina (expected average coverage will be several 100x). Do you have any idea how I could efficiently quantify my PCR products for pooling? (I thought about putting each product on a gel and pool different amounts of respective PCR products based on the brightness of the band on the gel. I'm not sure, however, whether this is accurate enough).

I'm happy for any suggestions!

[lorendarith]

(I thought about putting each product on a gel and pool different amounts of respective PCR products based on the brightness of the band on the gel. I'm not sure, however, whether this is accurate enough).

As you've said, it doesn't need to be perfect, pooling by band brightness works well enough. We did this for RAD-seq libraries and it turned out relatively balanced. However, instead of just pooling different amounts based on brightness, you could put equally bright bands into "groups" (not mixing them, just grouping) and then quantify a single representative from that group with a Qubit or whatever quantification method you prefer. This way you wouldn't have 1000 samples to measure, but maybe 10-20.

Otherwise, use a plate reader for quantification? This way you'd just have 10x 96 well plates, though the above option includes band visualization which you don't get with this quantification method.

[Marius]

Good input! Thank's!
If there are further ideas, please let me know!

[HESmith]

Why not qPCR? Three 384-well plates and you're done in less than a day.

[nucacidhunter]

After confirming success of PCR amplification, a high throughput approach would be to clean-up and concentrate (low elution volume) amplicons by using 0.5X AMPure beads. Then normalise a suitable portion of eluates with Normaliser beads (reaction can be scaled down). Normalised DNA can be pooled by volume by factoring in size to obtain equimolar representation of each amplicon in the pool. For instance 2 µl of 5 kb amplicons can be pooled with 4 µl of 10 kb amplicons.