TruSeq Cluster Generation

[TUCF JSS]

Hi All,

I've noticed that using the TruSeq genomic library prep kit, rather than my own stock of enzymes/materials, has yielded a much higher cluster density even when both samples are equimolar.

My theory is that the Illumina kit is working better than my own reagents, so the percent of successfully adapter-ligated DNA is higher, or somehow their proprietary blend of adapters work better than the ones I order from IDT. I do not have qPCR to quantify this, but I was just wondering if others had theories or noticed the same type of thing happening with their libraries.

Thanks!

[TonyBrooks]

We've seen the same thing. We've had a few TruSeq RNA libraries that have over clustered recently. I found your post after searching to see if anyone else had similar issues. We are thinking of switching the concentration we run all TruSeq libraries down to 11pM from our usual 13pM to compensate.

[pmiguel]

Concentration by qPCR? Or fluorimetry?

--
Phillip

[TonyBrooks]

We generally use fluorescence (QuBit) and the average size from the Bioanalyser. It seemed to work really well for the old paired end assay and any homebrew method.
We also qPCR'd some of the TruSeq 10nM stock that gave us overclustering and we found no correlation between which samples did over cluster and which seemed fine. There was some correlation between our fluorescence estimate and the qPCR though.

[protist]

Hello Tony,
We have been doing 'home brew' for a few years now and get good clusters but like you we have seen superior clustering with the TruSeq kit generated libraries. With the v5 kits and SCS2.9/RTA1.9 we have been loading 11-12 pM for our home brew libraries and getting fairly consistent cluster numbers (unless we use our indexed adapters ). We also quantify with Qubit. My suspicion is that the kit is way more efficient at the ligation step.

[pmiguel]

Probably some property of the secret oligos that Illumina includes in its "PPC" (PCR primer cocktail) of the TruSeq kit.

--
Phillip

[protist]

So very true! They are definitely a lot more secretive than they used to be before they sucked us all in!

[pmiguel]

Hello Tony,
We have been doing 'home brew' for a few years now and get good clusters but like you we have seen superior clustering with the TruSeq kit generated libraries. With the v5 kits and SCS2.9/RTA1.9 we have been loading 11-12 pM for our home brew libraries and getting fairly consistent cluster numbers (unless we use our indexed adapters ). We also quantify with Qubit. My suspicion is that the kit is way more efficient at the ligation step.

If it were just the ligation step, then Tony's qPCR assay would sort that out.
I think TruSeq amplicons may be more "flow-cell-trophic" and thus attach to the flow cell at higher titres than non-TruSeq amplicons.

--
Phillip

[pmiguel]

So very true! They are definitely a lot more secretive than they used to be before they sucked us all in!

Yes, very irritating. They really are treating the PPC oligos like a state secret. Ridiculous. I'm sure their competitors already know what they are (if they care.)

So the ones left in the dark are their customers.

I wonder if the granting agencies could just ban this sort of behavior among vendors selling to researchers funded by public monies?

--
Phillip

[TonyBrooks]

I vaguely remember Illumina mention that they use freakishly high thresholds for oligo quality in the TruSeq kits.....or maybe they were just using that as a reason why the kits are twice the price of homebrew.

[pmiguel]

Still -- run the PCR Primer Cocktail oligos on a lab chip -- they look ~80 nt long! So something is probably going on there.

--
Phillip

[ETHANol]

It is my guess, and this is purely a guess based on that fact that Phillip has seen that the oligos are much longer then what seems possible, is that the TruSeq PCR Primer Cocktail has 'overhanging sequence' which is probably extra binding sequences beyond what is in the adapters. If my TruSeq kit ever comes (I've been waiting over a month) I will Sanger sequence the amplicons to try to get to the bottom of this. If someone could mass spec them that would be nice!!! Alternatively they could contain LNA residues as well to enhance binding to the flow cell.

Illumina's kits are great, but for some applications we really do need to make the libraries with home-brew protocols, e.g. no TruSeq ChIP-seq kit yet, MeDIP-seq, 4C-seq, etc. And their delivery times to Greece doesn't mix well with the pace at which I like to do science.

[ETHANol]

By the way, TUFCS JSS you can check out my last couple runs which were done with you guys. Both had low cluster formation and were done with homemade primers similar to what is shown on your protocols. They were highly amplified libraries quantified by qPCR so the problem was not that there was material without adapters causing a high reading.

My guess to the cause:
1) Unused PCR primer in the library competing for binding to the flowcell. Now I'm using Ampure XP to purify the PCR product so this shouldn't be a problem with my next samples.
2) There is something different about Illumina's PCR primer cocktail (see above post). For now, I assume it is probably best to just understate the library concentration.

[ETHANol]

TUFC, one question, how many cluster were generated compared to how many you would get with TruSeq kits?

[LSAD]

Hello Tony,
We have been doing 'home brew' for a few years now and get good clusters but like you we have seen superior clustering with the TruSeq kit generated libraries. With the v5 kits and SCS2.9/RTA1.9 we have been loading 11-12 pM for our home brew libraries and getting fairly consistent cluster numbers (unless we use our indexed adapters ). We also quantify with Qubit. My suspicion is that the kit is way more efficient at the ligation step.

When we use our indexed primer, the cluster formation is not succesful.
Do you happen to know why it is like that?