SEQanswers

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



Similar Threads
Thread Thread Starter Forum Replies Last Post
error in GAII versus MiSeq versus HiSeq adaptivegenome Bioinformatics 3 11-10-2016 05:57 AM
What is the diff among Illumina Genome Analyzer, Miseq and hiseq? arkilis Illumina/Solexa 4 09-08-2013 04:24 PM
Custom recipe for GAII: one lane at a time mgravina Illumina/Solexa 0 03-27-2012 08:28 AM
Velvet Assembler: expected coverage versus estimated coverage versus effective covera DMCH Bioinformatics 1 11-30-2011 04:21 AM

Reply
 
Thread Tools
Old 01-14-2014, 04:32 PM   #1
ECO
--Site Admin--
 
Location: SF Bay Area, CA, USA

Join Date: Oct 2007
Posts: 1,358
Default MiSeq V2 versus V3 clustering...recipe diff

So I'm troubleshooting V2 vs. V3 clustering with the same exact samples (both of long and short inserts, all enriched gDNA libraries) and I wanted to see if anyone else has done any similar work.

See some recent data which shows generally half the cluster densities, as well as very poor performance of longer insert libraries on V3 kits.



So I went sleuthing on why this might be...suspecting changes to the clustering recipes....

Illumina calls out the only reagent differences are scan and incorporation mix, so I (*admittedly naively*) don't think they'd change the formulation of clustering reagents.

Quote:
Yes. There are two new reagents in MiSeq v3 kits:
  • New incorporation mix (IMT)
  • New scan mix (USM)
The PR2 volume is increased to 500 ml to support longer runs. In the near future, all MiSeq PR2 bottles will have a 500 ml fill to avoid confusion when starting a run. Look for announcements on the Illumina Tech Support Bulletin Board regarding this change.
I noted what I think are significant differences in clustering recipes (by looking at the recipe .xml's for "OnBoardClusterGeneration" steps, both are attached for the Default chemistry).

1. V2 chemistry pumps the template in first, then heats up to 75C, while V3 heats the FC up to 75C, then pumps the template in. (lines ~18-20 of the attached files).

2. The template rampdown step ("TMP RampDown") to 40C is twice the time for V3 than it is for V2.

3. "First Extension" is dramatically different as well. V3 is at 50C, followed by 6 cycles of pump-wait. V2 doesn't appear to change the temperature to 50C (so it's still at 40C from the Rampdown??), and then performs 15 cycles of pump-wait (with half the volume of reagents).

4. "Amplification1" is very different as well. V2 doesn't wait at all in between each reagent, but just waits for 15000 after the AMS1 is injected. V3 pumps each reagent (LDR, LPM, AMS1) and waits in between for 7200,7200,18700 respectively. Also V2 does 26 cycles, while V3 does 24.

5. "Linearisation" is similarly lengthened in V3 as in the previous step, adding waits in between reagent pumps (and cleaving for roughly 3x the time).

So I'm talking to Illumina TS, but I'm going to try running the V2 clustering recipe with a V3 kit, some of these mods seem significant that they could impact general clustering efficiency as well as long insert amplification. Any input would be appreciated, even if it's an Illumina employee telling me to stop reading recipe .xmls.
Attached Files
File Type: xml v2_clustering.xml (3.9 KB, 100 views)
File Type: xml v3_clustering.xml (5.5 KB, 54 views)
ECO is offline   Reply With Quote
Old 01-15-2014, 02:47 AM   #2
Vinz
Member
 
Location: Germany

Join Date: Dec 2010
Posts: 80
Default

Thanks! Very interesting thoughts.

Our impression is rather opposite of what you describe. We load 20% less on V3 and still get about 20% higher cluster densities. So in our hands V3 is more efficient in cluster formation (or registration).
This is based on amplicon sequencing with about 500bp product length. (>100 runs)
Vinz is offline   Reply With Quote
Old 01-15-2014, 08:31 AM   #3
ECO
--Site Admin--
 
Location: SF Bay Area, CA, USA

Join Date: Oct 2007
Posts: 1,358
Default

Ah that's the amplicon recipe though...time to diff those!
ECO is offline   Reply With Quote
Old 01-15-2014, 10:05 AM   #4
bnmtthws
Junior Member
 
Location: New York City

Join Date: Mar 2011
Posts: 4
Default

quick question from a relatively recent adopter: are you implying that the instrument is using different recipes to generate clusters depending on the library type/workflow specified in the sample sheet?

we've been performing a lot of low- to mid-diversity amplicon sequencing on our miseq, and are still trying to nail down the clustering on both v2 and v3 kits - since we work with a non-model organism and are used to doing the alignments and bioinformatics ourselves, so i've just been running everything designated 'fastq-only.' if there are different clustering recipes, are there flags we can set in the sample-sheet to ensure that it's using the recipe optimized for our sample type (amplicons?)

thanks for any insight!
bnmtthws is offline   Reply With Quote
Old 01-15-2014, 11:02 AM   #5
ECO
--Site Admin--
 
Location: SF Bay Area, CA, USA

Join Date: Oct 2007
Posts: 1,358
Default

Quote:
Originally Posted by bnmtthws View Post
quick question from a relatively recent adopter: are you implying that the instrument is using different recipes to generate clusters depending on the library type/workflow specified in the sample sheet?
I didn't mean to imply that, just that I needed to check it. I just confirmed that within a specific version, the clustering recipes are identical.

That is, for Default and Amplicon within V3 kits, the clustering recipe is identical.

So Vinz and I are seeing different phenomena. Will follow up when I have more data.
ECO is offline   Reply With Quote
Old 01-15-2014, 11:05 AM   #6
GenoMax
Senior Member
 
Location: East Coast USA

Join Date: Feb 2008
Posts: 7,077
Default

@ECO: We consistently see much better performance with V3 reagents than V2 (like Vinz).

What version of MCS are you using?
GenoMax is offline   Reply With Quote
Old 01-15-2014, 11:11 AM   #7
bnmtthws
Junior Member
 
Location: New York City

Join Date: Mar 2011
Posts: 4
Default

Quote:
Originally Posted by ECO View Post
I didn't mean to imply that, just that I needed to check it. I just confirmed that within a specific version, the clustering recipes are identical.

That is, for Default and Amplicon within V3 kits, the clustering recipe is identical.

So Vinz and I are seeing different phenomena. Will follow up when I have more data.
great - thanks for the clarification!
bnmtthws is offline   Reply With Quote
Old 04-12-2014, 09:14 AM   #8
Chit_HK
Junior Member
 
Location: Hong Kong

Join Date: Jul 2013
Posts: 3
Default

Hi ECO,
I also had similar experience on the Miseq V3. The library I used was Truseq Amplicon Cancer Panel. I only got ~400k/mm2 cluster density with V3, but I got higher density in V2 kit. Don't know why...
Do you have any idea?
Chit_HK is offline   Reply With Quote
Old 04-29-2014, 05:41 AM   #9
DRYTCYV
Member
 
Location: Portugal

Join Date: Apr 2011
Posts: 50
Default

Hi all,

Using the same procedure to prepare libraries (Nextera XT - no bead normalization) and loaded in V2 and V3 kits, we detected a higher cluster density in V3.
We are following the procedure to prepare a 4nM library->dilute in NaOH (5 min @room temp) and dilute the library to 15pM. the cluster density change from 1000-1100 to 1300-1400.

Do you have a worst R2 quality in V3 than V2?

Thanks in advance
DRYTCYV is offline   Reply With Quote
Old 05-06-2014, 07:46 AM   #10
DJ Flowcell
Member
 
Location: Boston

Join Date: Apr 2013
Posts: 13
Default

We found that the v3 kits can accommodate cluster densities from 1000 - 1200 K/mm^2. We load exactly the same for v2 and v3 kits. The improved chemistry of the v3 kits takes care of the higher cluster densities - at least that's what I found from my experience. We typically achieve high quality reads reaching ~15M for v2, and ~21M for v3.
DJ Flowcell is offline   Reply With Quote
Old 05-06-2014, 08:38 AM   #11
NextGenSeq
Senior Member
 
Location: USA

Join Date: Apr 2009
Posts: 482
Default

We've found that the higher density flowcells are very sensitive to proper denaturation.

I suspect that the high insert size samples aren't being denatured completely.

For non-amplicon libraries you must use molecular biology grade NaOH and it has to be fresh. Old solution stocks of NaOH will absorb CO2 and the pH will change. Seal your stock solutions of NaOH very well to prevent CO2 changing the pH.
NextGenSeq is offline   Reply With Quote
Old 05-06-2014, 08:58 AM   #12
DJ Flowcell
Member
 
Location: Boston

Join Date: Apr 2013
Posts: 13
Default

NextGenSeq brings up a good point. We've been told by Illumina technicians that the NaOH provided in their kits isn't necessarily the best and has been prone to the pH change over time as noted. They recommended buying 10N stocks that will better hold their pH over time when stored correctly. While I'm not too keen if the larger/smaller sized libraries require differences in their denaturing protocol, I do know for a fact that higher GC content libraries need longer denaturing times for full representation.
DJ Flowcell is offline   Reply With Quote
Old 05-06-2014, 11:35 AM   #13
pmiguel
Senior Member
 
Location: Purdue University, West Lafayette, Indiana

Join Date: Aug 2008
Posts: 2,317
Default

Anyone used the post-NaOH denaturation, heat denaturation method recommended by Illumina for "difficult templates".

BTW, at least in the past ECO wasn't using Illumina's crazy dilute the begezzus-out-of-it with buffer method of neutralization after denaturation. So that may be biting him here.
pmiguel is offline   Reply With Quote
Old 05-06-2014, 12:24 PM   #14
DJ Flowcell
Member
 
Location: Boston

Join Date: Apr 2013
Posts: 13
Default

Yes, we use the post-NaOH denaturation regularly for all our samples. We experience a wide array of libraries so we go ahead with the 2' incubation at 98 degrees after the NaOH denaturation at RT. It will absolutely help the libraries with high GC content and won't hurt the others so it might as well be done for all.
DJ Flowcell is offline   Reply With Quote
Old 05-08-2014, 05:18 AM   #15
pmiguel
Senior Member
 
Location: Purdue University, West Lafayette, Indiana

Join Date: Aug 2008
Posts: 2,317
Default

ECO? Don't leave us hanging...
What have you discovered subsequently?

--
Phillip
pmiguel is offline   Reply With Quote
Old 06-04-2014, 02:27 AM   #16
jen_is
Junior Member
 
Location: UK

Join Date: Jun 2014
Posts: 2
Default

Hello, this is the first I've heard of the post-NaOH denaturation, heat denaturation method. You say it is recommended by Illumina but is this documented anywhere? We have some high GC content libraries so it may help us.

Jen
jen_is is offline   Reply With Quote
Old 06-04-2014, 03:33 AM   #17
nucacidhunter
Jafar Jabbari
 
Location: Melbourne

Join Date: Jan 2013
Posts: 1,238
Default

It is documented in MyIllumina/Support Bulletins. You may need a MyIllumina account to log in and access it. The title is "Improved cluster density for difficult samples: the heat treatment" added on Apr 2 2014.
nucacidhunter is offline   Reply With Quote
Old 06-05-2014, 04:33 AM   #18
jen_is
Junior Member
 
Location: UK

Join Date: Jun 2014
Posts: 2
Default

That's great, I already have an illumina account, thanks!
jen_is 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 08:40 AM.


Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2020, vBulletin Solutions, Inc.
Single Sign On provided by vBSSO