Reducing the amount of DNA going into a TruSeq library Preparation?

[mchotalia]

Hi,

I was wondering if anyone has experience of using less than 1ug of input DNA for the TruSeq DNA library prep? I am assuming i would have to alter the covaris shearing settings and also dilute the adapters in the ligation step to avoid adapter dimers? Does anyone have tips for this?

In addition, i recently made TruSeq DNA libraries using 1ug of DNA and followed the instructions carefully. I ended up with so much library that i was thinking of next time quantifying the library after size selection and if there is enough then i would omit the enrichment step. Is there any reason why i should not do this?

It'll be great to hear your thoughts.

Thanks in advance.

[Heisman]

I haven't used TruSeq, but for your second question, you can have a lot of library that isn't necessarily properly prepped. So, without the PCR you won't be sure if your DNA has proper adapters. You can get around this issue by using a small portion of your DNA, doing PCR, and back calculating to determine how much properly prepped DNA you have (qPCR would work too, I think). I haven't tried this but I think it should work if your calculation is correct.

[pmiguel]

Possibly, but check this out:

The following:

Is there a PCR-free TruSeq Sample Prep protocol available?

The new index adapter design enables PCR-free protocols. (A single cycle of synthesis is required to separate the forked adapter.)

When I wrote to Illumina tech support asking if the NaOH used prior to clustering should not suffice for "separating" the "forked adapter" I got the following:

We have found that if you simply take the annealed fragment that has the forked adapters on each end and add it to the clustering reaction that even with NaOH treatment before clustering, there is poor initial hybridization and the resultant clustering is poor. By allowing at least one PCR cycle this separates the strands and the forked adapters, allows a single amplification of the entire strands, resulting in a clean, double-stranded species that gives cleaner clustering results.

We had a 70% GC genome organism, Deinococcus radiodurans, library that gave very patchy coverage with the normal 10 cycles of amplification and v1 chemistry, but when run with no amplification and v3 chemistry:

(1) The coverage was very smooth, but
(2) The library gave very few reads -- at 10x less than expected from qPCR.

Sadly we consumed the entire library before I stumbled upon the above information. So I can't verify that with a single round of amplification we would get the best of both worlds.

--
Phillip

[FredericRaymond]

Have you tried doing the PCR with enzymes, buffers or additives optimized for high-GC sequences? For example, reagents such as Betaine or commercial solutions can be added to the PCR to improve the melting of GCs.

FR

[pmiguel]

No, we just tried no amp. By qPCR there was plenty of library. It actually worked, but gave us far fewer clusters than we expected.

--
Phillip

[BIG_SNP]

We often go as low as 100ng input for DNA libraries. Alternatively as others have mentioned we have also had good luck with 1ug input and dramatically reducing the number of PCR cycles.

[mchotalia]

To Big SNP,

When you are using 100ng for the TruSeq kit do you alter the volume of adapters going into the reaction or do you continue as if you are using 1ug?

Many thanks

[pmiguel]

If you really have 100ng (ie, measured on a fluorimeter with a ds-specific fluor), then you are probably fine as you are. Your issue would be adapter dimers? Illumina built in two levels of adapter-dimer suppression:

(1) T-tailed adapters with nuclease resistant phosphorothioate linkages.
(2) PCR primers that include the terminal T (which won't be there if the T-tail was lost and allowing blunt end ligation of adapters.

Note well, however, (2) is speculative because Illumina refuses to reveal the sequence of their PCR primers! Probably just want to maintain some type of mystique to keep you interested in them...

--
Phillip

[csquared]

(1) T-tailed adapters with nuclease resistant phosphorothioate linkages.
(2) PCR primers that include the terminal T (which won't be there if the T-tail was lost and allowing blunt end ligation of adapters.

Note well, however, (2) is speculative because Illumina refuses to reveal the sequence of their PCR primers! Probably just want to maintain some type of mystique to keep you interested in them...

The TruSeq primers are well 5' of the ligation point so (2) is very very unlikely. From everything we can tell, there is nothing special about the primers and the sequences of the primers are basically the flow cell sequences at the 5' end of the TruSeq adaptors.

[pmiguel]

The TruSeq primers are well 5' of the ligation point so (2) is very very unlikely. From everything we can tell, there is nothing special about the primers and the sequences of the primers are basically the flow cell sequences at the 5' end of the TruSeq adaptors.

I don't think so.

Please see:

http://seqanswers.com/forums/showpost.php?p=48824&postcount=13

To me it looks like the TruSeq PPC (PCR Primer Cocktail) contains oligos that are about 80 nt.

But if you have a different explanation I would love to see it...

--
Phillip

[Heisman]

Phillip,

Is it possible that the primers have LNA's or some other non-normal entity that would cause them to run at a different length than their true length?

[csquared]

I don't think so.

To me it looks like the TruSeq PPC (PCR Primer Cocktail) contains oligos that are about 80 nt.

But if you have a different explanation I would love to see it...

Lots of things run at higher than expected sizes on the Bioanalyzer. If you are correct, the TruSeq primer cocktail is a mix of all 12 indexes and somehow the the different indexes don't cross-anneal. I'll stick with very very unlikely.

[pmiguel]

Lots of things run at higher than expected sizes on the Bioanalyzer. If you are correct, the TruSeq primer cocktail is a mix of all 12 indexes and somehow the the different indexes don't cross-anneal. I'll stick with very very unlikely.

Yep, if I am correct, that is one thing it might imply. Well, it does not imply that different indexes don't cross-anneal. In fact it suggests that there could be a certain amount of cross-annealing. When I wrote Illumina to point out that it would be potentially very bad if their primers included a mixture of indexes and begged them to basically "say it wasn't so", I got the following response:

Unfortunately our PCR primer sequences are proprietary and I am unable to comment on their size and/or composition.

Apart from that, it does seem pretty crazy and therefore unlikely. However, Agilient chips are designed to estimate the length of polynucleotides. Sure they are far from perfect. But I don't think I am being unreasonable in claiming that if the PCR primers run at 80-85 nt, that it is possible that they are, in fact, that length.

Your claim that "lots of things run at higher than expected sizes on the Bioanalyzer" is, well, absolutely the case. But not relevant here, I should think, because, all evidence of that sort I have seen indicates that ssDNA (or probably ssRNA) on a dsDNA chip runs slower than dsDNA of the same length. But the results I link to in my earlier post were from an RNA chip after heat denaturation of the "PCR Primer Cocktail". The sizing artifacts I see on RNA chips are of another sort and would not be expected to cause single stranded oligos to run larger than their actual length.

Anyway, my request for evidence that the PPC primers are not 80-85 nts long stands. But currently the only actual evidence I have seen indicates that they are that long.

Also, I should add, that even if they are 80-85 nts long, it does not mean they must include the indexes. Could be that they contain distal sequence not amplified during bridge PCR.

--
Phillip

[pmiguel]

Phillip,

Is it possible that the primers have LNA's or some other non-normal entity that would cause them to run at a different length than their true length?

Sure, that is possible. Not sure if LNA's do run longer than normal bases. But it could also be other modifications.

--
Phillip

[csquared]

Also, I should add, that even if they are 80-85 nts long, it does not mean they must include the indexes. Could be that they contain distal sequence not amplified during bridge PCR.

--
Phillip

Even less likely. Have a look at the flow cell sequences and how the bridge PCR works. Any 5' distal sequence would really be problematic.

I'm going to try to amplify a standard PE library with the TruSeq primers. If they amplify, we know that the primers are not full length. If they don't, maybe they are.

One final piece of information: I was able to get directly from Illumina (and a very good source) that the PCR primer cocktail does NOT contain a mix of all 12 indexes and that the sequences are not full length. They were not able to provide any more details than that.