SEQanswers

Go Back   SEQanswers > Bioinformatics > Bioinformatics



Similar Threads
Thread Thread Starter Forum Replies Last Post
RNA-Seq: GFOLD: a generalized fold change for ranking differentially expressed genes Newsbot! Literature Watch 8 03-07-2016 11:17 AM
A fold change heatmap for RNA seq analysis using CuffDiff and cummerbund devking RNA Sequencing 16 12-23-2014 09:34 AM
CPM in EdgeR - before and after exactTest Kaas RNA Sequencing 0 07-31-2013 04:00 AM
Cut-off values for fold change in RNA seq smalan RNA Sequencing 1 07-12-2012 10:10 AM
EdgeR Fold Change Calculation umnklang Bioinformatics 3 06-21-2012 12:17 AM

Reply
 
Thread Tools
Old 01-31-2014, 05:13 AM   #1
feralBiologist
Member
 
Location: UK

Join Date: Jun 2011
Posts: 61
Default edgeR: fold change reported by exactTest for zero values of rna-seq

I have used the exact test in edgeR to compute the log fold changes. Here is the snippet:

Code:
d <- DGEList(counts=counts, group=samples$Condition)
d <- calcNormFactors(d)
d <- estimateCommonDisp(d)
d <- estimateTagwiseDisp(d)
de <- exactTest(d)
I've noticed that some genes have zero expression in all samples belonging to one of the two conditions. This would make the fold change mathematically undefined (division by zero). Yet the FC is reported as being ~2^-9. My question is - how does edgeR come up with this value? I've checked both the manual and the reference guide but couldn't figure out. There are various functions that accept pseudocounts as parameters but I have entered none in my snippet. So how does edgeR make up for the zero values in this particular case (which seems to be the default usage of the exactTest)?
feralBiologist is offline   Reply With Quote
Old 01-31-2014, 06:33 AM   #2
rskr
Senior Member
 
Location: Santa Fe, NM

Join Date: Oct 2010
Posts: 250
Default

Good numerical analysts and mathematicians do delta epsilon proofs to figure out what a mathematically undefined quantity should be in specific cases to provide continuity, then redefine the definition in a specific instance. For example a correlation between two sets with zero variance, isn't defined, division by zero, but it's pretty obvious that a value of 1.0 or perfect correlation makes the most sense when doing hierarchical clustering.

A more general issue is that using fold changes is likely to amplify noise.
rskr is offline   Reply With Quote
Old 01-31-2014, 08:35 AM   #3
feralBiologist
Member
 
Location: UK

Join Date: Jun 2011
Posts: 61
Default

Thanks, rskr.

After closer inspection, exactTest seems to be using predFC function which, by default, adds a pseudocount of 0.125 to all observations. This seems to answer it. It would be, perhaps, more transparent to have this as a parameter in exactTest itself but once you dig in the documentation it becomes clear anyway.
feralBiologist is offline   Reply With Quote
Reply

Tags
edger, fold change, rna-seq

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 03:32 AM.


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