SEQanswers

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



Similar Threads
Thread Thread Starter Forum Replies Last Post
MiSeq run with low Q30 and high cluster PF haroldnunez Illumina/Solexa 12 01-06-2017 07:34 AM
Stopped MiSeq run memento Illumina/Solexa 24 10-28-2015 08:17 PM
MiSeq stopped run yaximik Illumina/Solexa 20 03-18-2013 08:10 AM
How much PhiX in MiSeq run? thdybwf Illumina/Solexa 1 11-26-2012 04:01 AM
Worst MiSeq Run ever? pmiguel Illumina/Solexa 0 04-30-2012 05:54 AM

Reply
 
Thread Tools
Old 02-17-2014, 09:08 PM   #1
ReGenMC
Junior Member
 
Location: San Diego

Join Date: Feb 2014
Posts: 1
Default MiSeq v3 2x300 run, Low Q30

I recently ran a ~400bp library prepped with NexteraXT on the MiSeq using reagents kit v3 for a 2x300 run. The project was DeNovo Assembly, paired end single index. I got high but in range cluster densities of 1383k/mm2 with 88.2% passing filter.

My issue is that the run only had 43% >Q30 bases. Does anyone have any ideas for troubleshooting?
ReGenMC is offline   Reply With Quote
Old 02-18-2014, 10:12 AM   #2
GW_OK
Senior Member
 
Location: Oklahoma

Join Date: Sep 2009
Posts: 383
Default

Perhaps it's because you ran into the adapter and then the nothingness beyond while doing a 400bp insert on a 2x300 run?
GW_OK is offline   Reply With Quote
Old 02-18-2014, 11:57 AM   #3
Wallysb01
Senior Member
 
Location: San Francisco, CA

Join Date: Feb 2011
Posts: 286
Default

Yeah, GW is probably right. In a ~400bp library, a lot of your fragments will be <300bp. If you trim adapters, what happens? Also, you can run something like Flash to overlap your reads, and it may give you a better sense of the size distribution of your library as it was sequenced.

Remember even if a bioanalyzer says your library is one size, clustering on the flowcell is not necissarily going to occure with the same efficiency across the various sizes in your library. And in general, smaller fragments cluster better. So if you see 400bp on the bioanalyzer, don't be at all surprised if after sequencing the mean size is closer to 300bp.
Wallysb01 is offline   Reply With Quote
Old 02-18-2014, 12:59 PM   #4
tbanks
Member
 
Location: St. Catharines

Join Date: Mar 2010
Posts: 11
Default

I just did two runs with the 600 cycle kit (2 x 300). One was 82%>Q30 and the other 81%. Average insert size of that library is 275 bp. 29 million reads passing filter for each run. I don't think it is your insert size.

Travis
tbanks is offline   Reply With Quote
Old 02-18-2014, 01:04 PM   #5
GenoMax
Senior Member
 
Location: East Coast USA

Join Date: Feb 2008
Posts: 6,579
Default

At almost 1400 clusters/mm^2 you are at the outer bounds of recommended cluster density for v.3 chemistry. It is not surprising that your Q30 values have suffered as a result. Short of loading less and re-running there is nothing you can do about the existing Q-scores.
GenoMax is offline   Reply With Quote
Old 02-18-2014, 01:42 PM   #6
nickloman
Senior Member
 
Location: Birmingham, UK

Join Date: Jul 2009
Posts: 356
Default

We've run much higher cluster densities and still stayed within the performance spec for V3 kits which is >70% Q30. First off, report this run to Illumina and get the kit replaced. If this problems recurs with other runs then you will need to debug further, but it could just be a faulty reagent batch.
nickloman is offline   Reply With Quote
Old 02-25-2014, 07:34 AM   #7
heidurlo
Junior Member
 
Location: Sweden

Join Date: Jan 2014
Posts: 2
Default

@nickloman
-How much higher cluster densities are we talking about?

I just ran a v3 600 cycle, cluster density around 1800k/mm2, and Q30 only 24%. The tech support told me that I had overclustering, which I believe, but I am just curious to know about your experience.
heidurlo is offline   Reply With Quote
Old 02-27-2014, 12:49 PM   #8
GenoMax
Senior Member
 
Location: East Coast USA

Join Date: Feb 2008
Posts: 6,579
Default

There is an implied aspect to @nickloman's comment. The value of cluster density is going to be somewhat dependent on the type and quality of the library e.g. if you are doing plain resequencing then you may be able to get away with higher densities as opposed to amplicon sequencing.

It is quite surprising how far one can push MiSeq limits and still keep getting good data. The cliff is precipitous and when you go over a limit the fall in quality may be similar to what you saw.
GenoMax is offline   Reply With Quote
Old 03-19-2014, 01:50 AM   #9
Riddhi
Junior Member
 
Location: Germany

Join Date: Feb 2014
Posts: 1
Default

Hi i ran the Miseq V3 150 PE cycles with 12pM dilution and i got 1981K cluster with Q30 77.3%

can any body guide me to reduce cluster and increase the Q30?
Should i lower my library dilution?
Riddhi is offline   Reply With Quote
Old 03-26-2014, 10:58 AM   #10
DJ Flowcell
Member
 
Location: Boston

Join Date: Apr 2013
Posts: 13
Default

Hi Riddhi,

Yes, definitely lower your cluster counts through a library dilution. 1900K/mm2 is very high. The V3 kits function optimally between 1000 - 1300. You could literally split your final library dilution in HT1 in half and have good cluster counts with good quality too.
DJ Flowcell is offline   Reply With Quote
Old 05-16-2014, 11:46 AM   #11
justinjun
Junior Member
 
Location: vancouver

Join Date: Feb 2014
Posts: 4
Default

Hi Riddhi,
I am surprised by the fact that 12pM can give you 1981K/mm2 cluster density.
I am a little bit doubting your DNA measurement. Do you use Qubit to do the measurement or what? I usually dilute my library to 14pM which, in turn, give me 1300~1400K density (I did run 3 times 2x300 run with different library with average DNA length of 600bp).
But, even with ~1400k density, Q30 is still pretty low (around 75%, though higher than kit performance spec.)
I am wondering how to increase the overall Q30? Any thought? Thanks.

Last edited by justinjun; 05-16-2014 at 11:49 AM. Reason: typo
justinjun is offline   Reply With Quote
Old 05-20-2014, 07:54 AM   #12
NextGenSeq
Senior Member
 
Location: USA

Join Date: Apr 2009
Posts: 482
Default

What type of library? We see lower quality scores with amplicon libraries.
NextGenSeq is offline   Reply With Quote
Old 05-20-2014, 12:55 PM   #13
justinjun
Junior Member
 
Location: vancouver

Join Date: Feb 2014
Posts: 4
Default

Quote:
Originally Posted by NextGenSeq View Post
What type of library? We see lower quality scores with amplicon libraries.
genomic DNA prepared using NEBNext Ultra DNA Library Prep Kit for Illumina, which has 10 cycle PCR amplification step.
justinjun is offline   Reply With Quote
Old 02-13-2015, 01:58 PM   #14
unladenedswallow
Junior Member
 
Location: Vancouver, BC

Join Date: Oct 2013
Posts: 2
Default

Our collaborators mentioned they had encountered a similar situation where their fragment sizes were too small and it had 'read into the nothing beyond'. However, it seems like other people on this forum were getting decent results.

I recently did a 2 x 300 run with 1220K/mm2 cluster density and our Q30% started dropping rapidly at cycle 250. We loaded several libraries of different average fragment sizes (~380 bp to ~900 bp) at equal volumes.

Do you guys find that 2x300 V3 runs give less consistent results? We just started running 600 cycle runs.

Quote:
Originally Posted by ReGenMC View Post
I recently ran a ~400bp library prepped with NexteraXT on the MiSeq using reagents kit v3 for a 2x300 run.
Quote:
Originally Posted by GW_OK View Post
Perhaps it's because you ran into the adapter and then the nothingness beyond while doing a 400bp insert on a 2x300 run?
unladenedswallow is offline   Reply With Quote
Old 02-13-2015, 03:20 PM   #15
Brian Bushnell
Super Moderator
 
Location: Walnut Creek, CA

Join Date: Jan 2014
Posts: 2,695
Default

How are you calculating the fragment sizes? The sizes observed by some tools prior to sequencing do not always correlate well with the the sizes calculated bioinformatically after sequencing.
Brian Bushnell is offline   Reply With Quote
Old 02-13-2015, 03:31 PM   #16
unladenedswallow
Junior Member
 
Location: Vancouver, BC

Join Date: Oct 2013
Posts: 2
Default

We use a Bioanalyzer to calculate average fragment size as recommended by Illumina.
unladenedswallow is offline   Reply With Quote
Old 02-13-2015, 04:51 PM   #17
Brian Bushnell
Super Moderator
 
Location: Walnut Creek, CA

Join Date: Jan 2014
Posts: 2,695
Default

I recommend you generate an insert-size histogram computationally, as well, to help determine whether this is a problem with chemistry, or with short inserts. You can do so without a reference with BBMerge, like this:
bbmerge.sh in=reads.fq ihist=ihist_merging.txt reads=1000000

...or if you do have an assembly, via mapping like this:
bbmap.sh in=reads.fq ref=reference.fa nodisk ihist=ihist_mapping.txt reads=1000000

The two approaches have different biases, and both will give somewhat different answers than Bioanalyzer.
Brian Bushnell is offline   Reply With Quote
Reply

Tags
2x300bp, low q30, miseq, reagent kit v3

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 12:17 AM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2017, vBulletin Solutions, Inc.
Single Sign On provided by vBSSO