Exogenous spike for RNA seq

veryconfused · 04-05-2017, 08:07 PM

Hi all... help! New to all this next generation sequencing stuff. I need to add an exogenous spike (have used c.elegans 54 for RT PCR) for serum samples for next generation sequencing. Cant spike in too much or not enough! I cant find any papers on how much spike to use (though i have learnt what a femtamole is!Lol) Can anyone help?
Thanks in advance!

cmbetts · 04-06-2017, 09:13 AM

The most commonly used exogenous spike in for RNA-Seq is the ERCCs sold through Thermo. Their user manuals have guidelines for adding the appropriate amount to different sample types to get the intended 1-5% of reads.

veryconfused · 04-06-2017, 04:10 PM

Thanks... i did look into the ERCC spike and spoke to the sequencing peeps and they mentioned ERCC spike ins were more to determine technical variations. I need an exogenous spike as i have no idea what my samples are going to do! I'm looking at RNA levels in cardiac surgery patients -so there are no "normal" or "healthy" groups i am able to use as a comparison for normal micro RNA levels... i have found a couple of papers that have used ath-miR-159a but, their methods aren't terribly clear and the corresponding author isn't answering my emails

Its proving a difficult task! But thanks for your input - i appreciate your reply

fanli · 04-07-2017, 06:34 AM

Another option:
http://www.sequin.xyz/about/rnaseq/

In my experience, ERCC spikes aren't very useful for normalization. See this paper:
http://www.nature.com/nbt/journal/v3.../nbt.2931.html

cmbetts · 04-07-2017, 11:50 AM

Are you sequencing mRNA or miRNA? I think it would be easier to give a suggestion if you described your experiment more.

I've never actually used the ERCCs for normalization. Most of the DE tools only want raw counts. I just use them to get a sense of assay performance (lots and lots of efficiency measurements for scRNA-Seq method dev).

veryconfused · 04-07-2017, 09:31 PM

I've asked the group at the Garvan Institute in Sydney about their sequins - but was advised against them by the group as there were "too complex" for my needs (not sure what that means!) The ERCC don't help as i need the exogenous spike in right at the beginning...
just a bit of background of my PhD - i'm looking at patients undergoing cardiac surgery with cardiopulmonary bypass (the machine that keeps patients alive while they are having open heart surgery - my job is managing this -(part-time PhD as well...insane, i know!)).
From my preliminary RT-PCR experiments a couple of microRNAs of interest went astronomically through the roof - hence the move to next gen sequencing to see what all the other microRNA's do. I have a pre operative sample and a during surgery sample but my supervisor is concerned i'll just get millions of reads in both samples and wont actually be able to see what is going on - hence the introduction of an exogenous spike (like miR-54 that i used for my RT-PCR for normalisation)... just really stuck on how much too add as there is not too much in the literature. Thanks again all

GenoMax · 04-08-2017, 05:25 AM

In your case it sounds like all you want/need is a technical spike to say that the reads for the spike are similarr in both samples leading to an assumption that the rest of the sample behaved identically (as far as the conversion to DNA/Sequencing went).

Someone will have to comment on what/how much to add but be aware that at the amounts we are talking about you will never get an identical number of reads in two samples for the spike.

If you have a clear signal from the tests so far it should be replicable, if it is real, across more samples. Having millions of reads should not mask it as long as you follow some established normalization procedure for your counts.

04-05-2017, 08:07 PM	#1
veryconfused Junior Member Location: australia Join Date: Apr 2017 Posts: 3	Exogenous spike for RNA seq Hi all... help! New to all this next generation sequencing stuff. I need to add an exogenous spike (have used c.elegans 54 for RT PCR) for serum samples for next generation sequencing. Cant spike in too much or not enough! I cant find any papers on how much spike to use (though i have learnt what a femtamole is!Lol) Can anyone help? Thanks in advance!

04-06-2017, 04:10 PM	#3
veryconfused Junior Member Location: australia Join Date: Apr 2017 Posts: 3	ERCC won't help :( Thanks... i did look into the ERCC spike and spoke to the sequencing peeps and they mentioned ERCC spike ins were more to determine technical variations. I need an exogenous spike as i have no idea what my samples are going to do! I'm looking at RNA levels in cardiac surgery patients -so there are no "normal" or "healthy" groups i am able to use as a comparison for normal micro RNA levels... i have found a couple of papers that have used ath-miR-159a but, their methods aren't terribly clear and the corresponding author isn't answering my emails Its proving a difficult task! But thanks for your input - i appreciate your reply

04-06-2017, 09:13 AM	#2
cmbetts Senior Member Location: Bay Area Join Date: Jun 2012 Posts: 121	The most commonly used exogenous spike in for RNA-Seq is the ERCCs sold through Thermo. Their user manuals have guidelines for adding the appropriate amount to different sample types to get the intended 1-5% of reads.

04-07-2017, 06:34 AM	#4
fanli Senior Member Location: California Join Date: Jul 2014 Posts: 198	Another option: http://www.sequin.xyz/about/rnaseq/ In my experience, ERCC spikes aren't very useful for normalization. See this paper: http://www.nature.com/nbt/journal/v3.../nbt.2931.html

04-07-2017, 11:50 AM	#5
cmbetts Senior Member Location: Bay Area Join Date: Jun 2012 Posts: 121	Are you sequencing mRNA or miRNA? I think it would be easier to give a suggestion if you described your experiment more. I've never actually used the ERCCs for normalization. Most of the DE tools only want raw counts. I just use them to get a sense of assay performance (lots and lots of efficiency measurements for scRNA-Seq method dev).

04-07-2017, 09:31 PM	#6
veryconfused Junior Member Location: australia Join Date: Apr 2017 Posts: 3	I've asked the group at the Garvan Institute in Sydney about their sequins - but was advised against them by the group as there were "too complex" for my needs (not sure what that means!) The ERCC don't help as i need the exogenous spike in right at the beginning... just a bit of background of my PhD - i'm looking at patients undergoing cardiac surgery with cardiopulmonary bypass (the machine that keeps patients alive while they are having open heart surgery - my job is managing this -(part-time PhD as well...insane, i know!)). From my preliminary RT-PCR experiments a couple of microRNAs of interest went astronomically through the roof - hence the move to next gen sequencing to see what all the other microRNA's do. I have a pre operative sample and a during surgery sample but my supervisor is concerned i'll just get millions of reads in both samples and wont actually be able to see what is going on - hence the introduction of an exogenous spike (like miR-54 that i used for my RT-PCR for normalisation)... just really stuck on how much too add as there is not too much in the literature. Thanks again all

04-08-2017, 05:25 AM	#7
GenoMax Senior Member Location: East Coast USA Join Date: Feb 2008 Posts: 7,091	In your case it sounds like all you want/need is a technical spike to say that the reads for the spike are similarr in both samples leading to an assumption that the rest of the sample behaved identically (as far as the conversion to DNA/Sequencing went). Someone will have to comment on what/how much to add but be aware that at the amounts we are talking about you will never get an identical number of reads in two samples for the spike. If you have a clear signal from the tests so far it should be replicable, if it is real, across more samples. Having millions of reads should not mask it as long as you follow some established normalization procedure for your counts.