Index Read distribution

RickC7 · 09-27-2017, 07:01 AM

Hello,

I'm looking for advice, help, thoughts, etc on how to achieve more balanced read output when multiplexing. Typically I run all libraries on Agilent Bioanalyzer and pool according to the library peak in equimolar amounts. My last TruSeq mRNA seq project had between 2% and 8% reads for each sample, which varied read output 12 to 48million reads per sample. I see variations also when doing 16s(Schloss barcodes). I just wondering if variation is typical, some barcodes may amplify better than others, etc?? Do most labs see more even read distribution or is it more likely to "do the best you can and hope the outcome is acceptable"?

I'd be interested in any success stories using normalization protocols, plates, beads etc for library pooling.

Thanks for any help!

nucacidhunter · 09-27-2017, 02:25 PM

The most accurate method for library quantification is qPCR.

RickC7 · 09-27-2017, 02:47 PM

Thanks, I get that and have no problem when loading a sequencer for optimal cluster density. I'm also not going to quantify 192 16s libraries via qPCR for multiple projects. I more looking to get better/ more even read distribution for multiplexed library pools.

nucacidhunter · 09-27-2017, 03:09 PM

For amplicons such as 16S that has been prepped using the same protocol, pooling based on dsDNA specific reagents such as PicoGreen results in up to 3x variation in output which is similar to bead or plate based normalisation methods.

DNA_Dan · 06-11-2018, 02:18 PM

Old post I know, sorry to be late to the party, but if you're still around RickC7, I'd like to see if you made any inroads on the normalization front. It seems like labs doing a lot of samples use the Nextera normalization method using beads and denaturing the library to get fragments off the bead. I am trying to create something similar for Truseq without denaturing.

RickC7 · 06-13-2018, 11:14 AM

I haven't yet tried bead based normalization, but I have tried some plate normalization methods like SequalPrep. It did not work well in our hands. For larger sample numbers, like when we are doing 16s, we now use fluoro-plate reader, then create a normalization plate template for our EPMotion. This works best, still get a few outliers but not bad. I randomly check a few samples to make sure unincorporated primers and such are removed.

For smaller projects <24 samples, I just use qubit/bioanalyzer. Calculate and dilute all libraries down to 4nM then pool. I get better read distribution and cluster density using qubit/bioanalyzer than qPCR. I can easliy squeeze out 600million PF reads on our Next-Seq at 85% PF and 85% Q30. My %Phix loading is usually close to being spot-on, so that tells me my quantification is accurate.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
How to get the median of read distribution in a genomic region?	jajclement	Bioinformatics	2	05-17-2016 01:53 PM
Read distribution at high sequence depth	ForeignMan	General	10	05-26-2011 04:50 AM
RNA-seq read distribution	wenhuang	RNA Sequencing	9	11-08-2010 06:00 PM
rna-seq read distribution	wenhuang	Bioinformatics	1	06-17-2010 10:07 AM
(a) Background Read Distribution (b) Spike-ins	ANJAN PURKAYASTHA	Bioinformatics	0	03-05-2009 08:02 AM

09-27-2017, 07:01 AM	#1
RickC7 Member Location: Baton Rouge, Louisiana Join Date: Feb 2010 Posts: 31	Index Read distribution Hello, I'm looking for advice, help, thoughts, etc on how to achieve more balanced read output when multiplexing. Typically I run all libraries on Agilent Bioanalyzer and pool according to the library peak in equimolar amounts. My last TruSeq mRNA seq project had between 2% and 8% reads for each sample, which varied read output 12 to 48million reads per sample. I see variations also when doing 16s(Schloss barcodes). I just wondering if variation is typical, some barcodes may amplify better than others, etc?? Do most labs see more even read distribution or is it more likely to "do the best you can and hope the outcome is acceptable"? I'd be interested in any success stories using normalization protocols, plates, beads etc for library pooling. Thanks for any help! Last edited by RickC7; 09-27-2017 at 08:06 AM.

09-27-2017, 02:25 PM	#2
nucacidhunter Jafar Jabbari Location: Melbourne Join Date: Jan 2013 Posts: 1,238	The most accurate method for library quantification is qPCR.

09-27-2017, 02:47 PM	#3
RickC7 Member Location: Baton Rouge, Louisiana Join Date: Feb 2010 Posts: 31	Thanks, I get that and have no problem when loading a sequencer for optimal cluster density. I'm also not going to quantify 192 16s libraries via qPCR for multiple projects. I more looking to get better/ more even read distribution for multiplexed library pools.

09-27-2017, 03:09 PM	#4
nucacidhunter Jafar Jabbari Location: Melbourne Join Date: Jan 2013 Posts: 1,238	For amplicons such as 16S that has been prepped using the same protocol, pooling based on dsDNA specific reagents such as PicoGreen results in up to 3x variation in output which is similar to bead or plate based normalisation methods.

06-11-2018, 02:18 PM	#5
DNA_Dan Member Location: Montana Join Date: Nov 2008 Posts: 21	Old post I know, sorry to be late to the party, but if you're still around RickC7, I'd like to see if you made any inroads on the normalization front. It seems like labs doing a lot of samples use the Nextera normalization method using beads and denaturing the library to get fragments off the bead. I am trying to create something similar for Truseq without denaturing.

06-13-2018, 11:14 AM	#6
RickC7 Member Location: Baton Rouge, Louisiana Join Date: Feb 2010 Posts: 31	I haven't yet tried bead based normalization, but I have tried some plate normalization methods like SequalPrep. It did not work well in our hands. For larger sample numbers, like when we are doing 16s, we now use fluoro-plate reader, then create a normalization plate template for our EPMotion. This works best, still get a few outliers but not bad. I randomly check a few samples to make sure unincorporated primers and such are removed. For smaller projects <24 samples, I just use qubit/bioanalyzer. Calculate and dilute all libraries down to 4nM then pool. I get better read distribution and cluster density using qubit/bioanalyzer than qPCR. I can easliy squeeze out 600million PF reads on our Next-Seq at 85% PF and 85% Q30. My %Phix loading is usually close to being spot-on, so that tells me my quantification is accurate.