Downsample data for MiSeq Reporter

kmcarr · 01-28-2013, 06:44 AM

As the amount of raw data generated by the MiSeq continues to grow we are starting to run into situations where even this "Benchtop" sequencer produces more data than is needed (or wanted) for a de novo genome assembly of things like bacteria and fungi, even with the Micro and Nano options. If I am going to do a de novo assembly offline I can simply downsample the data myself. But what if I want to use the built in MiSeq Reporter de novo assembly pipeline? Is there any way to configure it to only use a portion of the raw data for assembly?

kcchan · 01-28-2013, 09:27 AM

Currently the default assembly pipeline in the MiSeq can only handle 550Mb of sequence. You can decrease it by playing with the config files in MSR. You can read the MSR Theory of Operations document available on the Illumina website.

However, it seems strange to be throwing perfectly good reads away. Have you considered multiplexing more samples into the run so you generate more useful data?

kmcarr · 01-28-2013, 09:32 AM

Quote:

Originally Posted by kcchan

Currently the default assembly pipeline in the MiSeq can only handle 550Mb of sequence. You can decrease it by playing with the config files in MSR. You can read the MSR Theory of Operations document available on the Illumina website.

However, it seems strange to be throwing perfectly good reads away. Have you considered multiplexing more samples into the run so you generate more useful data?

Thanks for the info. Regarding multiplexing, yes that would be ideal, but as a core we have to take what researchers send us. Sometimes all they have is one sample.

GenoMax · 01-28-2013, 09:45 AM

This could be a good opportunity to spread some goodwill around. You can donate the spare capacity to deserving labs on campus, who are truly unable to afford the cost.

genbio64 · 01-28-2013, 11:05 AM

Is this feature new in the last software update? Illumina had told us there was virtually no flexibility when it came to the downstream pipelines through MSR.

GenoMax · 01-28-2013, 12:10 PM

Quote:

Originally Posted by genbio64

Is this feature new in the last software update? Illumina had told us there was virtually no flexibility when it came to the downstream pipelines through MSR.

MiSeq reporter v.2.1.43 was released just last week so it may be a new feature or an undocumented one Illumina wants people to generally stay away from.

kcchan · 01-28-2013, 12:27 PM

It's definitely good practice not to mess with the default analysis settings. Changing the settings will affect all future analysis, so know what you're doing before playing with the config files and revert them back if the next person needs the default settings. Most of the time I find it easier to do the downstream analysis myself and not on the instrument.

mcnelson.phd · 01-29-2013, 04:16 AM

Can I ask why you would even want Reporter to do de novo assembly? Unless things have changed with the newest version, Reporter (and Basespace) does no trimming or quality discrimination during its assembly, and I can't remember if it properly handles paired data. Every assembly done using Reporter that I've seen looks worthless compared to one's that I've done using Velvet on my own.

I guess if people are just sending you samples to sequence and intend to assemble it themselves, and you give them the assembly that Reporter spits out as a "bonus" along with the raw data then there's no real harm. But I think relying solely on an assembly done using Reporter or Basespace is pretty sloppy.

My opinion only of course though.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Do not use "adapter trimming" in MiSeq Reporter 2.0.25	ECO	Illumina/Solexa	9	12-10-2013 04:38 AM
alignment of Miseq data	mathew	Bioinformatics	3	01-07-2013 03:32 AM
Who uses a MiSeq for 16S data?	capsicum	Metagenomics	0	11-21-2012 01:10 PM
De-Multiplexing MiSeq Data	abh	RNA Sequencing	2	11-09-2012 04:26 AM
Should I trim my MiSeq data?	cwzkevin	Bioinformatics	3	09-06-2012 09:51 AM

01-28-2013, 06:44 AM	#1
kmcarr Senior Member Location: USA, Midwest Join Date: May 2008 Posts: 1,178	Downsample data for MiSeq Reporter As the amount of raw data generated by the MiSeq continues to grow we are starting to run into situations where even this "Benchtop" sequencer produces more data than is needed (or wanted) for a de novo genome assembly of things like bacteria and fungi, even with the Micro and Nano options. If I am going to do a de novo assembly offline I can simply downsample the data myself. But what if I want to use the built in MiSeq Reporter de novo assembly pipeline? Is there any way to configure it to only use a portion of the raw data for assembly?

01-28-2013, 09:27 AM	#2
kcchan Senior Member Location: USA Join Date: Jul 2012 Posts: 186	Currently the default assembly pipeline in the MiSeq can only handle 550Mb of sequence. You can decrease it by playing with the config files in MSR. You can read the MSR Theory of Operations document available on the Illumina website. However, it seems strange to be throwing perfectly good reads away. Have you considered multiplexing more samples into the run so you generate more useful data?

01-28-2013, 09:45 AM	#4
GenoMax Senior Member Location: East Coast USA Join Date: Feb 2008 Posts: 7,091	This could be a good opportunity to spread some goodwill around. You can donate the spare capacity to deserving labs on campus, who are truly unable to afford the cost.

01-28-2013, 11:05 AM	#5
genbio64 Member Location: New York Join Date: Dec 2009 Posts: 42	Is this feature new in the last software update? Illumina had told us there was virtually no flexibility when it came to the downstream pipelines through MSR.

01-28-2013, 12:27 PM	#7
kcchan Senior Member Location: USA Join Date: Jul 2012 Posts: 186	It's definitely good practice not to mess with the default analysis settings. Changing the settings will affect all future analysis, so know what you're doing before playing with the config files and revert them back if the next person needs the default settings. Most of the time I find it easier to do the downstream analysis myself and not on the instrument.

01-29-2013, 04:16 AM	#8
mcnelson.phd Senior Member Location: Connecticut Join Date: Jul 2011 Posts: 162	Can I ask why you would even want Reporter to do de novo assembly? Unless things have changed with the newest version, Reporter (and Basespace) does no trimming or quality discrimination during its assembly, and I can't remember if it properly handles paired data. Every assembly done using Reporter that I've seen looks worthless compared to one's that I've done using Velvet on my own. I guess if people are just sending you samples to sequence and intend to assemble it themselves, and you give them the assembly that Reporter spits out as a "bonus" along with the raw data then there's no real harm. But I think relying solely on an assembly done using Reporter or Basespace is pretty sloppy. My opinion only of course though.