SEQanswers

Go Back   SEQanswers > Applications Forums > Sample Prep / Library Generation



Similar Threads
Thread Thread Starter Forum Replies Last Post
Illumina DNA sequence specific strand bias involving orphan reads Tally Bioinformatics 1 06-21-2012 05:26 AM
Looking up specific position in bam Ooinp Bioinformatics 2 12-09-2011 01:58 PM
pileup only on a specific position foxyg Bioinformatics 8 12-22-2010 05:46 PM

Reply
 
Thread Tools
Old 10-18-2012, 05:27 AM   #1
Rocketknight
Member
 
Location: Ireland

Join Date: Sep 2011
Posts: 86
Default Position-specific GC bias?

Hi, I've been having an unusual problem with my reads. I've been doing exome sequencing on a HiSeq 2000, but I'm getting only about 40% GC content at the start of each read, climbing to 45-50% closer to the centre. At first I thought this was due to PCR bias. Changing my PCR protocols helped a bit, but there's still a very strong positional effect which I don't think can be explained by PCR alone. Has anyone seen something like this before?

We're using a Bioruptor for shearing as we don't have access to a Covaris. Is it possible that the Bioruptor is preferentially shearing in AT-rich regions? Is there any way to avoid this?

Rocketknight is offline   Reply With Quote
Old 10-18-2012, 09:04 AM   #2
HESmith
Senior Member
 
Location: Bethesda MD

Join Date: Oct 2009
Posts: 498
Default

We use the Bioruptor for gDNA libraries and never see a positional bias, so it's unlikely to be the source. My money is on enrichment as the cause.
HESmith is offline   Reply With Quote
Old 10-18-2012, 09:52 AM   #3
protist
Senior Member
 
Location: Ireland

Join Date: Jan 2009
Posts: 101
Default

We too have used a Biorupter for many gDNA libraries including DNAs from organisms with high GC content and have not seen any positional bias. I agree that the enrichment step may be the problem area
protist is offline   Reply With Quote
Old 10-18-2012, 09:57 AM   #4
Hamid
Senior Member
 
Location: San Diego, CA

Join Date: Sep 2009
Posts: 105
Default

Hi Rocketknight,

While the the enrichment step is a possible culprit, there is some merit to the shearing methodology used generating a bias in the library preparation. traditional sonicator unfortunately lack the thermal control required, so during shearing, they introduce quite a bit of heat the samples. This heat will melt high AT containing regions due to their low melting point, making them more susceptible to shearing as compared to the higher GC content regions.
Since you are using a bath sonicator for shearing to ~200bp, the processing time is quite long (>40minutes). These are also often carried out in plastic tubes which absorb acoustic energy and transmit it as heat to your sample. The plastic tubes then act as an insulator preventing the rapid dissipation of heat from your samples.
I have attached some thermal and pressure profiles of different sonciation methods and their effects on samples which gives a great idea of the thermal exposure of your DNA samples during processing using different sonication methods.

Thank you

hamid
Attached Files
File Type: pdf Energy and Temperature profile.pdf (440.0 KB, 19 views)
Hamid is offline   Reply With Quote
Old 10-18-2012, 01:06 PM   #5
HESmith
Senior Member
 
Location: Bethesda MD

Join Date: Oct 2009
Posts: 498
Default

Hi Hamid,

Having used the Covaris as well as the Bioruptor, I appreciate the relative advantages of each (I'm a particular fan of the tight size distributions offered by Covaris). However, I'd like to clarify a few points:

1) The Bioruptor uses a circulating water bath, much like the Covaris, so heating is not as much of an issue as you state.
2) Processing time for shearing to 200bp is 15 minutes of sonication (30 minutes @ 50% duty cycle), not >40 minutes.
3) Like protist, we have used the Bioruptor to construct gDNA libraries from a variety of organisms with different G/C contents and have not observed a positional bias in any of them. If you have data to the contrary, that would be very useful information to know.

-Harold
HESmith is offline   Reply With Quote
Old 10-19-2012, 10:55 AM   #6
Hamid
Senior Member
 
Location: San Diego, CA

Join Date: Sep 2009
Posts: 105
Default

Hi Harold,

I do work for Covaris as a scientist, but my post on seqanswers are not sales or marketing related as I do not have an interest in that capacity. I do have an interest in the science behind sample preparation for NGS, and ChIP, and its effect on endpoint analysis.

You are absolutely correct that bath sonicators also have recirculating baths to control the temperature during processing. The only drawback is that the circulation stops during processing, and is only active during the rest periods. Even that considered, as the pressure and temperature profile data I posted show, the samples and the water bath heat up during processing. it is an unfortunately side effect of uncontrolled bath and probe sonicators that has been described in literature. In a bath sonicator you are exposing the entire water bath to sonication, so heating is unavoidable.
with regards to processing time on bath sonicators, I have seen a very wide range 0f 30-60 minutes depending on the researcher and their DNA samples. with AFA for a 200bp shearing the total processing time is only 3 minutes regardless of sample concentration. Optimization has not been necessary due to sample concentration differences.
Hamid is offline   Reply With Quote
Old 10-19-2012, 02:16 PM   #7
HESmith
Senior Member
 
Location: Bethesda MD

Join Date: Oct 2009
Posts: 498
Default

Quote:
Originally Posted by Hamid View Post
Hi Harold,
I do have an interest in the science behind sample preparation for NGS, and ChIP, and its effect on endpoint analysis.
But the effect on our data (and that of protist) appears to be negligible. Would you point me to data that do exhibit an effect? I, too, am interested in potential sources of bias introduced during sample preparation.

Thanks,
Harold
HESmith is offline   Reply With Quote
Old 10-20-2012, 09:22 AM   #8
lorendarith
Guest
 

Posts: n/a
Default

I haven't look if there is a positional "GC bias" until now, but here you have one of our standard PE libraries, 150bp GAIIx and 100bp HiSeq2000.
Library prep is the same, the additional difference is the shearing to a desired insert size. The 150bp lib was sheared to 260bp and the 100bp lib to 460bp.

Both of them have this little peak on the 6th base. I don't know why the % is more uniform after that for the 150bp compared to the 100bp reads.
Attached Images
File Type: png per_base_gc_content.png (8.6 KB, 10 views)
File Type: png per_bas_gc_content.png (10.0 KB, 5 views)
  Reply With Quote
Reply

Tags
bias, bioruptor, covaris, gc content, illumina

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 12:36 AM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2018, vBulletin Solutions, Inc.
Single Sign On provided by vBSSO