Sciclone automation and Nextera Rapid Exomes

[LTJensen]

We are in the process of implementing a Sciclone robot to prep Nextera Rapid exomes. However, test samples fail the tagmentation step. Using the robot work flow we get tagment sizes well below 200bp (test1 attached).
The robot workflow differs from the Illumina protocol in the mixing procedure (robot mixes by pipetting, while Illumina protocol recommends shaking for 1min @ 1800rpm) and 58oC incubation (robot incubates in PCR plate, while manual suggests Midi plate).
We have tried to let the robot add buffer and enzyme, then perform off-robot shaking at 1800rpm in PCR plate, then returning plate to robot for incubation and adding of stop buffer (mixing by pipetting). This increased the tagment size to approximately 220bp (test 5 attached). However, after PCR, the profiles still look bad (test 5 NLS). Same approach with 1800rpm mixing and off robot incubation using Midi plates instead gave the same result.
Does anyone have any experience using the Sciclone for the Nextera Rapid protocol and/or have any ideas on how to resolve these problems?
Thanks!

[nucacidhunter]

From detailed information supplied here, robot does not seem to be the problem. The reason that Illumina recommends MIDI plate not PCR plate is that tagmenation reaction and later bead used for cleaning up the reaction cannot be mixed by shaking if they are in a PCR plate. MIDI plate wells have higher volume and are wider making solution in them suitable for mixing by a shaker. MIDI plate does not fit in PCR machines and they recommend Microheating system because it has insert for MIDI plate to fit. So MIDI plates are recommended for high throughput processing of samples.

Tagmentation reactions traces indicate that problem could be with DNA quality. If you are following Illumina recommended QC and input DNA amount, I would suspect that only some portions of input DNA are accessible for transposon so they get over-tagmented resulting in smaller fragments and some parts are not so accessible making that large DNA fragment trail stretching beyond upper marker at 10 kb. I would clean input DNA with AMPure beads or Zymo gDNA column and elute it with Tris-Cl pH8-8.5 or TE0.1 followed by QC and tagmentation and see if that makes difference. Close examination of DNA extraction method might be informative as well. We follow Illumina protocol in PCR plate because we do not process many samples at a time and the results are as illustrated in Illumina user guide.

[LTJensen]

Thank you for your reply, nuracidhunter.

I am pretty sure that DNA quality is not an issue here, as the used sample are high quality Promega reference DNA, purified by UF filtration. The same Promega sample produces fine tagmentation and post PCR Agilent profiles when prepared manually...

[cfrazar]

I'm currently running the same protocol on the Sciclone with 96 samples. I took a look at your BA traces and they look just like ours. I ran ~30 samples and they all had the same profile. When overlaid the leading edge of the peak was consistent among samples. It is likely due to ampure cleanup removing smaller fragments.

Small insert sizes tend to be a problem with Nextera. Initial DNA quantification is pretty key and plays a role in the resulting fragment size. You are using Promega DNA, so then quality should be pretty good. I would continue your test and go through the captures. If you need to, you can play with the final SPRI cleanup to push the average insert size up.

Good luck.

[GenoMax]

For anyone interested in a similar protocol using Tecan instead: http://res.illumina.com/documents/pr...automation.pdf

[LTJensen]

Thank you very much for the inputs.
cfrazar, how did the samples perform when sequenced?