[Abha]

Low Enrichments using Lib A EMPCR kits with SV emulsions

We are seeing similar kind of problems with our emPCR. The enrichment are very low even after increasing the CPB from 1CPB to 8 CPB! The enrichments still remains under 3%. We quantify our libraries using the Kapa QPCR kits which has been working very well previously and we used 1CPB or max 2CPB to get perfect enrichment. But for the last three FLX runs, the enrichment is considerably low even at higher cpb and the readlength have dropped from average 400bp to 250bp.
Still debating if it is the libraries, thermocycler or the emPCR kit.

[pmiguel]

Has anyone tried running the qPCR or emPCR products on a gel or an Agilent Bioanalyzer chip? Might shed some light...

Could be problems with small amplicon contamination.

Alas, Roche makes it nearly impossible for the end user to trouble shoot issues of this sort because they so stringently filter low quality sequence reads. So, really, you have no choice but to lay siege to the (generally overburdened) tech support apparatus at Roche.

(If anyone knows the right parameters to use to completely turn off filtering in the analysis pipeline, please share! I have gotten to the point where I have turned off everything I know of, but still the majority of bad reads are filtered. They just get caught by the "short reads" filter instead of an earlier one.)

--
Phillip

[Abha]

yes we ran the QPCR products on the gel, and we see nice clean single bands, as these were amplicon products and don't expect any small fragments to be around as they have been double cleaned up with Ampure XP

[pmiguel]

How about a little of the emPCR aqueous phase? To be clear, we have never tested this, but since it represents the liquid phase of the emPCR reaction, it might give some clues.

I wonder if using the new GS-FLX+ emPCR reaction conditions would help? It should not be necessary, but if Roche has been tinkering with the kits in preparation for GS-FLX+, there may have been some negative fall out for non-"plus" protocols.

--
Phillip

[yaximik]

So, what is the end result of this interesting discussion? Did you figure out what was the problem, or as often happens it just diasappeared leaving no clues?

[ajthomas]

Well, in our case (which started the whole discussion), it was indeed a problem with the library prep kits. I'm still not sure how it happened, but several kits all went bad at the same time, even kits we got a few months apart (same lot number, but different orders). We never figured out for sure which enzyme was the problem, but it was something in the library prep kit.

[Abha]

I don't think the problem has been resolved yet, the enrichments certainly have improved since that time. however now it seems like it is a RL adaptor issue. In our latest run of XLR70 of the 8 Lanes, we had perfect reads on 4 lanes with between 30- 40Mb of data per lane and read lengths of 400plus. however, the other 4 lanes failed miserably at the key pass stage itself ( got only 50k key passes per lane). And the difference between the samples which passed or failed were, the passed lanes had libraries prepared using the MID tag adaptors from May last year and the failed lanes had libraries prepared from standard rapid library adaptors from recently ordered kits. Roche is investigating this issue at this stage.
Should mention that the enrichments from all the libraies used were perfect with the expected range. I believe the problem is either with the sequencig primer or the sequencing binding site of the adaptor. Waiting to get some feedback from roche regarding this.