SEQanswers

Go Back   SEQanswers > Sequencing Technologies/Companies > Illumina/Solexa



Similar Threads
Thread Thread Starter Forum Replies Last Post
Library PCR enrichment questions (with image!) ZAAB Sample Prep / Library Generation 6 06-07-2011 07:47 PM
TruSeq Sample Prep, Enrichment PCR primer sequences EdibleMetal Sample Prep / Library Generation 11 04-13-2011 12:15 PM
Number of PCR How many PCR cycles to enrich adapter-modified DNA fragments MGH Man Sample Prep / Library Generation 5 07-26-2010 06:15 AM
PCR enrichment of libraries in 12 cycles or less? seqgirl123 Illumina/Solexa 6 07-05-2009 03:53 PM
exon enrichment by PCR for resequencing vasvale Genomic Resequencing 2 11-07-2008 02:17 PM

Reply
 
Thread Tools
Old 09-11-2008, 03:50 PM   #1
barb
Junior Member
 
Location: USA

Join Date: Sep 2008
Posts: 2
Default Adapter Enrichment PCR

Hi Solexa community,

Has anybody had the problem of having two bands after the enrichment PCR? The second band is bigger than the band I anticipated to see. Any ideas about what could have caused this?
__________________
Barbara
barb is offline   Reply With Quote
Old 09-24-2008, 11:46 AM   #2
cjohns
Junior Member
 
Location: Seattle

Join Date: Sep 2008
Posts: 7
Default

If it is about 120bp it is adapter complexes. We had this problem, but solved it by modifiying the size selection part of the protocol.
cjohns is offline   Reply With Quote
Old 09-24-2008, 04:13 PM   #3
barb
Junior Member
 
Location: USA

Join Date: Sep 2008
Posts: 2
Default

No, it is not at 120bp. My expected bands were at ~250, ~300 and ~350 (I cut several sizes) and then respectively a second band was present at ~400, ~500 and ~633. In the end I just did 3 enrichment pcr reactions, pooled,purified, ran it on a gel, extracted and purified the correct band. I retained enough DNA to submit that sample for sequencing.
__________________
Barbara
barb is offline   Reply With Quote
Old 10-08-2008, 11:57 AM   #4
tomv
Junior Member
 
Location: Cambridge MA

Join Date: Aug 2008
Posts: 8
Default

Quote:
Originally Posted by cjohns View Post
If it is about 120bp it is adapter complexes. We had this problem, but solved it by modifiying the size selection part of the protocol.
Can you elaborate on this. We sometimes get a high percentage of "linker" reads in our samples but its not clear to me how they could be forming. If I understand the protocol correctly the linkers should not be able to self anneal. So how does this happen and what does this "adapter complex" look like?
tomv is offline   Reply With Quote
Old 10-08-2008, 01:23 PM   #5
cjohns
Junior Member
 
Location: Seattle

Join Date: Sep 2008
Posts: 7
Default

Hi tomv,

We know that in our first round of sequencing we got a huge amount of product that gave us a spike on our bioanalyzer read (prior to sequencing) that runs at approx 120bp. We didn't know what it was, so we went ahead and put the samples on the Solexa. When we looked at the %base call plots we saw spikes that when one 'reads' them across they have the same sequence as: GATCGGAAGAGCTCGTATGCCGTCTTCTGCTTG which is one of Illumina's adapters (what I think you are calling a linker).

On the protocol that UCSC genome center has: http://www.genomecenter.ucdavis.edu/...brary_prep.pdf
they discuss an 'adapter-adapter band' that migrates at 120bp and they mention that cutting out this band during size selection will get rid of this.

I discussed this with one of my colleagues who believes that the adapters are concatenating during ligation. This starts with the two adapters linking together forming chains of different sizes. Then during the amplification step, the chain that is 3 adapters long is most efficiently amplified creating a large amount of 124 bp product.

We solved this problem by increasing the amount of ChIP DNA template and by running a longer agarose gel at slower speeds during the size selection part so that the 'adapter complexes' are separated away from the desired DNA pieces. Since doing this, we have not had this problem again.
cjohns is offline   Reply With Quote
Old 10-08-2008, 01:37 PM   #6
tomv
Junior Member
 
Location: Cambridge MA

Join Date: Aug 2008
Posts: 8
Default

Thanks cjons,

This is consistent with what we have seen. When we look at the full 36 base consensus sequence for these adapter reads we actually get the full adaptor sequence, plus the first 4 bases of the adapter again indicating that there are likely at least two copies of that adaptor in the final template.

Interestingly, the two adapter sequences in the read are separated by an "A". I'm sure this is significant in the ligation but I'm not sure how. This is the part that is most puzzling to me. I love to hear if anyone can explain it before I just put it in the old “weird stuff happens” category.

In any case, thanks for the tip about gel purification. We'll see if that helps.
tomv is offline   Reply With Quote
Old 10-06-2009, 11:43 AM   #7
Hbseq
Junior Member
 
Location: Boston,MA, USA

Join Date: Sep 2009
Posts: 2
Default 2nd larger DNA band after enrichment PCR

Quote:
Originally Posted by barb View Post
No, it is not at 120bp. My expected bands were at ~250, ~300 and ~350 (I cut several sizes) and then respectively a second band was present at ~400, ~500 and ~633. In the end I just did 3 enrichment pcr reactions, pooled,purified, ran it on a gel, extracted and purified the correct band. I retained enough DNA to submit that sample for sequencing.
Hi all, I'm new. I have encountered a similar problem as Barb posted before. Basically after enrichment PCR, I got 2 bands on agarose gel. One band is about the expected size, the other is of larger size, usually ~200bp bigger. I've tried different ways but can't solve the problem yet. Any help is appreciated! -Hbseq
Hbseq is offline   Reply With Quote
Old 10-07-2009, 03:07 AM   #8
andibody
Member
 
Location: Vienna

Join Date: Mar 2008
Posts: 45
Default

Hi,
We have observed this second band or peak too.
Also it is double the size of the expected one. Cutting during size selection at 200 bp will result in two peaks one at 200bp the second at 400bp indicating the formation of duplicates during PCR.
Surprisingly it doesn't happen each sample prep. Maybe a sample specific problem?
andibody is offline   Reply With Quote
Old 10-07-2009, 04:27 AM   #9
elaney_k
Member
 
Location: Dublin, Ireland

Join Date: Mar 2008
Posts: 55
Default

Barb, Hbseq & Andibody,

How many cycles are you using for your enrichment PCR? I used to see this using genomic DNA samples with the old library prep protocols that suggested ~18 cycles for the enrichment PCR. I haven't seen the larger band since I dropped the enrichment cycle number to 10-12.

Elaine
elaney_k is offline   Reply With Quote
Old 10-07-2009, 05:36 AM   #10
andibody
Member
 
Location: Vienna

Join Date: Mar 2008
Posts: 45
Default

I do ChiP-Seq + 18 cycles PCR. Reducing the number of cycles might be a good idea as we have also already seen that we end up anyhow being out of exponential enrichment when performing 18 cycles.
We'll give it a try.
Thanks
andibody is offline   Reply With Quote
Old 10-07-2009, 07:09 AM   #11
Hbseq
Junior Member
 
Location: Boston,MA, USA

Join Date: Sep 2009
Posts: 2
Default

Thanks for the responses. It doesn't appeared to be sample specific on my side. However, reducing PCR cycles or DNA template amount helped to reduce the intensity of the larger MW DNA band. Increasing annealing temperature had the reverse effect. I used 18 cycle PCR for enrichment.
Hbseq is offline   Reply With Quote
Old 10-07-2009, 12:07 PM   #12
cnicolet
Member
 
Location: Los Angeles

Join Date: Dec 2008
Posts: 35
Default

Hi:
I think the larger band is some kind of single strand product, possibly generated from asymmetric PCR at the last library prep step (our main reason for this conclusion is that we can generate them by heating a library that has run as a single band previously). Sometimes we see strong ones, sometimes we don't and it doesn't always correlate with cycle number. Sometimes the products are laddered, but sometimes they show up as very high molecular weight smears. Agilent claims the bioanalyzer dye won't recognize SS, but either there's residual reactivity or the single strand forms are "conformers' of some kind with intramolecular base pairing. The main problem we find with these is that they mess up the quantitation of the library--I'm pretty sure they will form sequencable clusters, but is their fluorescence reflective of the amount that's there? We don't think they are weird chimeras that would mess up an assembly, but don't have direct data for that.

Charlie
cnicolet is offline   Reply With Quote
Old 10-07-2009, 03:22 PM   #13
mollusc
Junior Member
 
Location: cambridge

Join Date: Oct 2009
Posts: 9
Default

our main reason for this conclusion is that we can generate them by heating a library that has run as a single band previously. Agilent claims the bioanalyzer dye won't recognize SS, but either there's residual reactivity or the single strand forms are "conformers' of some kind with intramolecular base pairing.

Hi Cnicolet,
Even i see this prob. and i suspect it to be ssDNA,buti have got no basis to confirm.Could you please elaborate how you concluded this and also about the conformers thing you mentioned about.
Thanks a lot
mollusc is offline   Reply With Quote
Old 10-07-2009, 03:56 PM   #14
cnicolet
Member
 
Location: Los Angeles

Join Date: Dec 2008
Posts: 35
Default

Hi Mollusc:

A couple of things, first, regarding cycle number--we thought it was overamplification at first but then a comparison of 10 vs 14 cycles showed the exact same intensity ratio between the main band and the faint upper MW band, it's just that both bands were more intense with 14 cycles. So while amplification probably plays a role, it's not the whole story we don't think. My evidence for SS is mostly indirect. As I said, my best evidence is when I heat a library that shows a perfect single band DS library of correct MW and run that, I see a laddered set of high MW bands generated from that library. I start to see it at 65, then it's all I see at 80 --the presumed DS band has completely gone but the laddered bands remain. It also "looks" like SS to me--not a sharp band (or peak, I'm just so used to gels) but more dispersed, fuzzy material that runs higher (like the difference between M13 ss and RF from the OLD days of sequencing). Also, you can imagine that the Tm of the two primers are very different because of the differences in length, so it doesn't seem a stretch to imagine asymmetric PCR occurring. Because intramolecular annealing should happen pretty fast, and we know the ends of the adapters at least are inverted repeats, I again feel it's possible that SS libraries would form hairpins of various kinds that would bind dye but migrate aberrantly.
cnicolet is offline   Reply With Quote
Old 10-08-2009, 03:14 AM   #15
elaney_k
Member
 
Location: Dublin, Ireland

Join Date: Mar 2008
Posts: 55
Default

These guys seem to have come across the same problem: http://seqanswers.com/forums/showthread.php?t=2582

I have also performed a second gel-isolation of some of my samples post enrichment PCR when I've had adapter dimers in my samples (extra band ~110bp) which worked well for me so maybe a second gel isolation step is an easy next step?
elaney_k is offline   Reply With Quote
Old 10-08-2009, 03:30 PM   #16
mollusc
Junior Member
 
Location: cambridge

Join Date: Oct 2009
Posts: 9
Default

I again feel it's possible that SS libraries would form hairpins of various kinds that would bind dye but migrate aberrantly.[/QUOTE]


Thanks for your helpCnicolet!

I always had this question.What are the factors controlling the formation of secondary structures of ssDNA.Is it just the repetitive sequences or anything else as well?
mollusc is offline   Reply With Quote
Old 10-27-2009, 03:54 AM   #17
sulfobus
Member
 
Location: Denmark

Join Date: Oct 2009
Posts: 12
Default 2nd larger DNA band after enrichment PCR

I frequently see this phenomenon, not only realted to Illumina sequencing. I believe it is mostly heteroduplexes, i.e. partially ssDNA that partially hybridize to other strands with a segment of similar sequence. This will form bulges of ssDNA that will slow gel migration, and can explain how it could appear just by denaturing/renaturing the DNA.

It is also seen in Quail et al, Nature Methods, 5:12 (2008). As stated in that article and in this and other threads, limiting the number of PCR cycles or limit the DNA load will decrease the formation. It is formed during the plateau phase of the PCR.

Other PCR artefacts take place in the PCR plataue, most notably recombinations, so there are many reasons not to dwell in the plateau for too long.
sulfobus is offline   Reply With Quote
Old 01-20-2010, 06:51 AM   #18
tomv
Junior Member
 
Location: Cambridge MA

Join Date: Aug 2008
Posts: 8
Default

We have an interesting twist on this issue. We have had samples that have been gel purified post-PCR and then, with nothing else happening to the samples, when we run them on a HS-BioA chip we see TWO peaks with the second peak ~2x the size of the gel fragment. This would seem to rule out the asymetric PCR theory, as those ssDNA products should have been removed in the gel. Could it be that the dsPCR products are "recombineing" at the adaptor sequences to form heterodiamers?
tomv is offline   Reply With Quote
Old 01-21-2010, 06:37 AM   #19
andibody
Member
 
Location: Vienna

Join Date: Mar 2008
Posts: 45
Default

When we saw those double bands in summer last year we started to have additional problems, especially a reduced efficacy of adapter ligation and a decrease of unique reads. We started using NEB sample prep reagents, using only paired end adapters (also for single read runs), aliquoting dATPs in minimal amounts, doing a-ligation and adapter-ligation very carefully (never stopping in between these two, vials always on ice, etc.) and never had again any of mentioned problems including double bands. What exactly resolved the problems, I can't tell, not even if the problems we saw were directly related to one another but I suppose the solution was one of that modifications to the protocol. This is a highly anecdotal story, sorry, but I thought it might still help someone.
protocol: 5ng starting material, adaptors 1:20, pre-PCR gel purificacion,18 cycles PCR.
andibody is offline   Reply With Quote
Old 04-27-2010, 12:34 PM   #20
HCMVSeq
Junior Member
 
Location: USA

Join Date: Feb 2010
Posts: 1
Default Same problems with PCR enrichment

Hello Everyone,

I've had the exact same problem with my Illumina PCR enrichment step prior to sequencing...I've had two different-sized PCR products show up...one that is my expected size (about 250-300bp) and another that is about double that (500-600bp). I've done a PCR "cycle titration" to see when each band shows up during PCR. The attached image "PCR Illumina 3" shows my initial PCR product after 17 cycles. The attached image "PCR Illumina 2" shows a cycle titration every 5 cycles from 10 to 35. As you can see, cycle 15 shows the lower (expected product), but the remaining cycles show just the larger product. Finally, "PCR Illumina" image shows a finer cycle titration (every cycle from 11 through 18 except 14...forgot to take sample ...) and you can clearly see a shift from the lower MW product to the higher MW product. My solution is to do a gel extraction of the lower MW band at cycle 15 (most lower MW product...most likely still in the exponential phase of PCR)...I chose to do a gel extraction to remove any primers and possible "primer/adapter concatemers" that have been discussed in this thread and others (didn't see any on gel, but I guess it's better to be safe than sorry). I'm not sure what this shift in MW means...but I like the idea of ssDNA secondary structures that has been mentioned in this thread.

Good luck everyone,
Pete
Attached Images
File Type: jpg PCR Illumina 3.jpg (33.6 KB, 145 views)
File Type: jpg PCR Illumina 2.jpg (25.0 KB, 154 views)
File Type: jpg PCR Illumina.jpg (18.1 KB, 176 views)
HCMVSeq is offline   Reply With Quote
Reply

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off




All times are GMT -8. The time now is 11:26 PM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2021, vBulletin Solutions, Inc.
Single Sign On provided by vBSSO