Primer contamination in library

Morten · 10-16-2014, 02:44 AM

Hi all,

We have been preparing libraries for Illumina sequencing using the ScriptSeq v2 kit. This gave us a bit of trouble and the resulting library is not really perfect. There is a large peak at 69 bp.
Please see the bioanalyzer image.

The peak at 69 bp is about the size of a single primer. Not an adapter-dimer.
My question is would this be a problem for cluster generation and sequencing?

This type of contamination would not be amplified, so I guess it would not be a problem, but I might be missing something? Would you feel safe sequencing this library?

Thank you,
Morten.

tegudet · 10-16-2014, 03:05 AM

Hi Morten,
I think this product could be sequenced and contaminate your reads.
I suggest you to use BluePippin, where you select the size of the product you want (let say 90bp to 400bp), so that you exclude 69bp primer contamination.

Good luck,

tegudet

Morten · 10-16-2014, 04:22 AM

Thank you for your reply tegudet.
However, the yield of this library is not good at all. There is simply not enough material to attempt a size selection.
As I see it my options are either sequence as is or remake the library.

microgirl123 · 10-16-2014, 06:58 AM

We sequence libraries with excess adapter on the MiSeq all the time. It's the adapter dimers that cause the problem. Single adapters can bind to the flow cell, but cannot bridge amplify so are not sequenced. Adapter dimers both bind and bridge amplify so they cause significant problems.

Morten · 10-17-2014, 04:30 AM

Thanks a lot for the first hand info!
I will go ahead and have the library sequenced.

Morgane_AUS · 12-02-2014, 03:44 PM

Quote:

Originally Posted by Morten

Thanks a lot for the first hand info!
I will go ahead and have the library sequenced.

How did it go ?

Morgane_AUS · 12-02-2014, 03:53 PM

Quote:

Originally Posted by microgirl123

We sequence libraries with excess adapter on the MiSeq all the time. It's the adapter dimers that cause the problem. Single adapters can bind to the flow cell, but cannot bridge amplify so are not sequenced. Adapter dimers both bind and bridge amplify so they cause significant problems.

Hi,

Have you ever attempted to sequence a library with such a high proportion of adapter dimers ? Please see the gel here:
https://bioinfo4biologists.wordpress...on-of-the-day/

We planned to use a single end 50 bp read (MiSeq V2 - 15 M reads, 24 samples, genome size 4.41 Mbp). Basically the cell might be flogged with adapters. My reasoning is as follow

Cartridge: 15 Million reads
Targeted number of reads: 10 Million reads
25% is ChiPed DNA: 2.5 Million reads
24 Samples: 0.1 Million reads/sample
50 bp per read: 5.2 Million bp /sample ---> 1x genome coverage ->BAD

Any hint would be greatly appreciated.

Morgane

cmbetts · 12-02-2014, 05:10 PM

Are you sure that's actually adapter dimer and not a primer-dimer o the library amplification primers? The full Illumina adapters are about twice the size of your peak (~120bp, assuming these are TruSeq compatible), and primer dimers essentially invisible to the sequencer since they don't have all of the necessary flowcell and priming sequences. I do see a little of the 120bp peak, but I'd expect it to be <5% of reads...

Morgane_AUS · 12-02-2014, 05:56 PM

Quote:

Originally Posted by cmbetts

Are you sure that's actually adapter dimer and not a primer-dimer o the library amplification primers? The full Illumina adapters are about twice the size of your peak (~120bp, assuming these are TruSeq compatible), and primer dimers essentially invisible to the sequencer since they don't have all of the necessary flowcell and priming sequences. I do see a little of the 120bp peak, but I'd expect it to be <5% of reads...

Mmm.. maybe you're right.. NEBNext ultra DNA adaptor are about 70 bp long and form a hairpin that would produce a ~35 bp double stranded fragment.

The primers though are 60 bp and could be responsible of the 120 bp peak.

Now what I'm wondering is whether the adaptors can make concatemers ?

Are you saying that you would sequence a library looking like this ?

Thanks !

cmbetts · 12-02-2014, 06:22 PM

I guess it depends on how precious your sample is. That's a slightly higher MW peak than I'm used to seeing for library amplification primers (but I've also nevr used this particular kit), and while I wouldn't expect them to sequence, if they are true primer-dimers and not excess primers, they might still have some of the necessary sequences to interfere with proper sequencing of the library. I'd also be concerned about that super-wide library size distribution as well. It would be difficult to know just how much to load on the sequencer since I'd expect that a sizable amount of it is too large to successfully cluster. I'd still be tempted to try a two-sided Ampure size selection and elute in a small volume to concentrate.

Morgane_AUS · 12-02-2014, 06:36 PM

Quote:

Originally Posted by cmbetts

I guess it depends on how precious your sample is. That's a slightly higher MW peak than I'm used to seeing for library amplification primers (but I've also nevr used this particular kit), and while I wouldn't expect them to sequence, if they are true primer-dimers and not excess primers, they might still have some of the necessary sequences to interfere with proper sequencing of the library. I'd also be concerned about that super-wide library size distribution as well. It would be difficult to know just how much to load on the sequencer since I'd expect that a sizable amount of it is too large to successfully cluster. I'd still be tempted to try a two-sided Ampure size selection and elute in a small volume to concentrate.

I didn't do the two sided size selection because the protocol didn't recommend to do it when starting the lib prep with less than 100 ng (and I started with 0.25 ng !). I'm worried I will loose the DNA if I do it now. These samples are quite precious (results of working the last 5 week-ends - it is a side project) and I won't be able to repeat the whole experiment due to other time constraints.

I was thinking about doing one bead purification step to remove the 120 bp fragment. Do you think large fragments can affect the whole sequencing run?

Thanks for you advice !

nucacidhunter · 12-02-2014, 09:00 PM

Quote:

Originally Posted by Morgane_AUS

Hi,

Have you ever attempted to sequence a library with such a high proportion of adapter dimers ? Please see the gel there: https://bioinfo4biologists.wordpress...dapter-dimers/

Could you attach your figure. I cannot open this link.

Morgane_AUS · 12-02-2014, 09:29 PM

Quote:

Originally Posted by nucacidhunter

Could you attach your figure. I cannot open this link.

sorry I don't know what's wrong with the link.

Here it is. Normal means that ChIPed DNA was used directly as input of the lib prep. WGA means ChIPed DNA was amplified using a whole genome amplification kit before lib prep. I started the lib prep with 0.25 ng of ChIPed DNA and 2.5 ng of amplified ChIPed DNA.

Morgane_AUS · 12-02-2014, 09:39 PM

updated link
https://bioinfo4biologists.wordpress...on-of-the-day/

nucacidhunter · 12-02-2014, 09:50 PM

It is difficult to figure out the nature of band in 100-200 bp mostly because of overloaded ladder. A TapeStation or Bioanalyser would have been better for diagnosis. If you had a no-template library prep that also would have helped with working out the possible source of that band. In any case, you can clean it with 0.9x bead twice. Larger fragments won’t an issue as cluster generation favours smaller fragments.

Morgane_AUS · 12-02-2014, 09:56 PM

We have the Qsep (similar to bioanalyser) but something went wrong with the cartridge so I had to use a gel. I had to overexpose it in order to see the smear (there is only 1 uL of ladder..).

I will include a no template library prep next time, I couldn't here because of the number of samples I have and kit size.

may I ask why you would do the 0.9x bead twice ?

Thanks for your insights !

nucacidhunter · 12-02-2014, 10:02 PM

One time wash less likely will get rid of that band. If a small residue of that band (prsumably primer- or adapter-dimer) is left in youer sample, they will form clusters more efficiently and you may get very small portion of reads from your target fragments.

Morgane_AUS · 12-02-2014, 10:04 PM

I don't have much DNA in there so I'm worried I will loose so much by doing it twice... I guess there is no win win situation here ! Finger crossed !

nucacidhunter · 12-02-2014, 10:08 PM

I do not know how many PCR cycles you have done. But you can do 2-3 cycles after 1st clean up.

cmbetts · 12-03-2014, 09:29 AM

I was about to suggest the same thing. Post-PCR, you always have the option of hitting it with a few more cycles if you end up with low yield. At this point, all of the fragments have been amplified enough that you don't really need to worry about drop-out anymore, you just run the risk of further distorting their relative abundance.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Library peak-let contamination at high end of TruSeq BioAnalyzer trace	Acacianpunk	Sample Prep / Library Generation	3	06-10-2014 07:13 PM
Primer dimer in ChIP-seq library	lavagirl	Sample Prep / Library Generation	7	06-07-2013 05:52 AM
Newbler Warning - Primer contamination	loba17	Bioinformatics	6	06-19-2012 05:47 AM
454Jr run metrics - possible PCR primer contamination	NGS4YetiB	454 Pyrosequencing	0	08-18-2011 02:54 PM
Plasmid contamination in Long Tag Paired End library	kmcarr	454 Pyrosequencing	3	03-11-2009 05:29 PM

10-16-2014, 03:05 AM	#2
tegudet Junior Member Location: Tübingen, Germany Join Date: Oct 2014 Posts: 1	Hi Morten, I think this product could be sequenced and contaminate your reads. I suggest you to use BluePippin, where you select the size of the product you want (let say 90bp to 400bp), so that you exclude 69bp primer contamination. Good luck, tegudet

10-16-2014, 04:22 AM	#3
Morten Junior Member Location: Denmark Join Date: Jun 2010 Posts: 9	Thank you for your reply tegudet. However, the yield of this library is not good at all. There is simply not enough material to attempt a size selection. As I see it my options are either sequence as is or remake the library.

10-16-2014, 06:58 AM	#4
microgirl123 Senior Member Location: New England Join Date: Jun 2012 Posts: 200	We sequence libraries with excess adapter on the MiSeq all the time. It's the adapter dimers that cause the problem. Single adapters can bind to the flow cell, but cannot bridge amplify so are not sequenced. Adapter dimers both bind and bridge amplify so they cause significant problems.

10-17-2014, 04:30 AM	#5
Morten Junior Member Location: Denmark Join Date: Jun 2010 Posts: 9	Thanks a lot for the first hand info! I will go ahead and have the library sequenced.

12-02-2014, 05:10 PM	#8
cmbetts Senior Member Location: Bay Area Join Date: Jun 2012 Posts: 118	Are you sure that's actually adapter dimer and not a primer-dimer o the library amplification primers? The full Illumina adapters are about twice the size of your peak (~120bp, assuming these are TruSeq compatible), and primer dimers essentially invisible to the sequencer since they don't have all of the necessary flowcell and priming sequences. I do see a little of the 120bp peak, but I'd expect it to be <5% of reads...

12-02-2014, 06:22 PM	#10
cmbetts Senior Member Location: Bay Area Join Date: Jun 2012 Posts: 118	I guess it depends on how precious your sample is. That's a slightly higher MW peak than I'm used to seeing for library amplification primers (but I've also nevr used this particular kit), and while I wouldn't expect them to sequence, if they are true primer-dimers and not excess primers, they might still have some of the necessary sequences to interfere with proper sequencing of the library. I'd also be concerned about that super-wide library size distribution as well. It would be difficult to know just how much to load on the sequencer since I'd expect that a sizable amount of it is too large to successfully cluster. I'd still be tempted to try a two-sided Ampure size selection and elute in a small volume to concentrate.

12-02-2014, 09:39 PM	#14
Morgane_AUS Member Location: Australia Join Date: Aug 2014 Posts: 25	updated link https://bioinfo4biologists.wordpress...on-of-the-day/

12-02-2014, 09:50 PM	#15
nucacidhunter Jafar Jabbari Location: Melbourne Join Date: Jan 2013 Posts: 1,238	It is difficult to figure out the nature of band in 100-200 bp mostly because of overloaded ladder. A TapeStation or Bioanalyser would have been better for diagnosis. If you had a no-template library prep that also would have helped with working out the possible source of that band. In any case, you can clean it with 0.9x bead twice. Larger fragments won’t an issue as cluster generation favours smaller fragments.

12-02-2014, 09:56 PM	#16
Morgane_AUS Member Location: Australia Join Date: Aug 2014 Posts: 25	We have the Qsep (similar to bioanalyser) but something went wrong with the cartridge so I had to use a gel. I had to overexpose it in order to see the smear (there is only 1 uL of ladder..). I will include a no template library prep next time, I couldn't here because of the number of samples I have and kit size. may I ask why you would do the 0.9x bead twice ? Thanks for your insights !

12-02-2014, 10:02 PM	#17
nucacidhunter Jafar Jabbari Location: Melbourne Join Date: Jan 2013 Posts: 1,238	One time wash less likely will get rid of that band. If a small residue of that band (prsumably primer- or adapter-dimer) is left in youer sample, they will form clusters more efficiently and you may get very small portion of reads from your target fragments.

12-02-2014, 10:04 PM	#18
Morgane_AUS Member Location: Australia Join Date: Aug 2014 Posts: 25	I don't have much DNA in there so I'm worried I will loose so much by doing it twice... I guess there is no win win situation here ! Finger crossed !

12-02-2014, 10:08 PM	#19
nucacidhunter Jafar Jabbari Location: Melbourne Join Date: Jan 2013 Posts: 1,238	I do not know how many PCR cycles you have done. But you can do 2-3 cycles after 1st clean up.

12-03-2014, 09:29 AM	#20
cmbetts Senior Member Location: Bay Area Join Date: Jun 2012 Posts: 118	I was about to suggest the same thing. Post-PCR, you always have the option of hitting it with a few more cycles if you end up with low yield. At this point, all of the fragments have been amplified enough that you don't really need to worry about drop-out anymore, you just run the risk of further distorting their relative abundance.