SEQanswers

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



Similar Threads
Thread Thread Starter Forum Replies Last Post
Hiseq max cluster density? fffhiseq Illumina/Solexa 19 04-03-2012 05:41 AM
Are Illumina missing a trick over cluster density? henry.wood Illumina/Solexa 13 01-13-2011 01:18 PM
library concentration and cluster density cybog337 Illumina/Solexa 0 01-12-2011 02:44 AM
Cluster density Illumina CatVincent Introductions 2 11-23-2010 10:50 AM
Factors affecting cluster density SeqTruth Illumina/Solexa 3 11-09-2010 06:59 AM

Reply
 
Thread Tools
Old 10-19-2009, 07:13 AM   #1
andibody
Member
 
Location: Vienna

Join Date: Mar 2008
Posts: 45
Default Troubles with Cluster Density

Hi all,

We have had sporadic troubles with cluster density. Our samples usually yield between 150.000 and 210.000 colonies per tile (GA IIx).
Last weeks flowcell showed only ~90.000 colonies. Samples originated from different preparations making it unlikely them being the cause.
Our proposed suspects: cluster gen kit / flowcell / sample denaturation.
Has anybody had similar problems or knows which the critical steps in cluster generation are?
We check concentration with bioanalyzer after sample prep; didn't include PhiX.


Thanks
andibody is offline   Reply With Quote
Old 10-19-2009, 04:03 PM   #2
mollusc
Junior Member
 
Location: cambridge

Join Date: Oct 2009
Posts: 9
Default

we always aim for 210K clusters which is pretty easily acheived by quantifying the library on QPCR.The numbers are reproducible.variations in the cluster numbers might be because of pipetting errors.We had this problem intially, but overcame this prob by weighing the hyb buffer when making template hybs.May i ask you how much would you load onto flowcell to hit this number?
Thanks!
mollusc is offline   Reply With Quote
Old 10-29-2009, 04:10 AM   #3
andibody
Member
 
Location: Vienna

Join Date: Mar 2008
Posts: 45
Default

Sorry for the late reply. We use a 6.5 pM but know that we could go a little bit higher. Which protocol do you use for QPCR quantification? At which step do you wheig hyb buffer? Before mixing with NaOH and subsequently with sample?
The strange thing is we are used to a certain variance but from time to time it drops significantly without obvious reason...
andibody is offline   Reply With Quote
Old 10-30-2009, 04:53 AM   #4
elaney_k
Member
 
Location: Dublin, Ireland

Join Date: Mar 2008
Posts: 55
Default

Hi Andibody,

What type of preps are you doing? I've seen it with Chip-seq and mRNA-seq preps were further analysis of the sequenced reads showed a percentage of adapter-dimers in the samples. This happened when the input was lower than expected for the protocol. QPCR showed up to 3 ct difference between samples even though they were all quantified and diluted to the same concentration (10nM).
So whilst pipetting errors are definitely something to consider, any large variation in cluster number that I've come across has been due more to the sample prep where low input amounts (or low yield of polyA RNA during RNA-seq prep) leads to adapter-dimers in your final library. I would check the sequences that came off the run for presence of adapters.

elaine
elaney_k is offline   Reply With Quote
Old 11-03-2009, 02:50 AM   #5
andibody
Member
 
Location: Vienna

Join Date: Mar 2008
Posts: 45
Default

Hi Elaine,

Most of our samples are indeed ChIP-Seq or RNA-Seq and in many cases starting amount is less than 10ng.
We are at the moment setting up the QPCR hoping to get more reliable concentration estimations.
We have checked in the past for adaptor contamination but haven't got many hits.
How do you use Adaptor sequences as template for alignment? As single sequence or concatenated or randomly extended to the read length?

Best
Andi
andibody is offline   Reply With Quote
Old 11-03-2009, 03:49 AM   #6
elaney_k
Member
 
Location: Dublin, Ireland

Join Date: Mar 2008
Posts: 55
Default

Hi Andi,

I look for a slightly truncated adapter sequence in the export files using the grep command so for ChIP-seq (single read adapters) the command I use is:
grep -c "GATCGGAAGAGCTCGTATGCCGTCTTCT" s_N_export.txt
and for RNA-seq (paired-read adapters) the command is:
grep -c "GATCGGAAGAGCGGTTCAGCAGGAAT" s_N_export.txt

The majority of our RNA-seq libraries have a 0 count for the adapter sequence but the odd few, where the final library yield was <5ng/ul resulted in much lower than expected cluster numbers on the flowcell and correlated with detection of adapter sequences in the sequenced reads.

Elaine
elaney_k is offline   Reply With Quote
Old 11-03-2009, 04:52 AM   #7
andibody
Member
 
Location: Vienna

Join Date: Mar 2008
Posts: 45
Default

I used Bowtie for alignment to adaptor sequence and even allowed for 2 mismatches but haven't got hits so I suppose that this is not the cause for my reduced cluster density. I hope that the QPCR will give more insight into this...


Best
Andi
andibody is offline   Reply With Quote
Old 11-03-2009, 08:16 AM   #8
elaney_k
Member
 
Location: Dublin, Ireland

Join Date: Mar 2008
Posts: 55
Default

Hi Andi,

Does Bowtie allow you to align reads to a reference that's shorter than your read length?
elaney_k is offline   Reply With Quote
Old 11-03-2009, 08:38 AM   #9
andibody
Member
 
Location: Vienna

Join Date: Mar 2008
Posts: 45
Default

Hi Elaine,

No it doesn't, thanks for that hint.
I'll try out your grep command.
Why didn't you use the last 4 bases of the adaptor sequence?
andibody is offline   Reply With Quote
Old 11-03-2009, 10:28 AM   #10
elaney_k
Member
 
Location: Dublin, Ireland

Join Date: Mar 2008
Posts: 55
Default

Hi Andi,

I was initially looking through data from a ChIP-seq run that had a really low % alignment and by eye I could see the same sequence repeated, when i checked the sequence it turned out to be a slightly shorter version of the adapter sequence. Less than 1% of the reads contained the full adapter sequence but up to 20% contained the truncated form.
I tried both full length adapter sequence and a slightly shorter version with the RNA-seq data when the cluster number dropped dramatically and found a much higher percentage of the shorter adapter sequence in the reads than the full-length again. The sequence wasn't just at the end of the reads either. I'm sure there's a logical molecular biology explanation for the phenomenon but I don't know exactly what it is myself.

Elaine
elaney_k 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 01:43 PM.


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