[BioC] edgeR - expressen between genes

[Jenny Drnevich]

Hi all,

>Comparing expression b/w genes (as opposed to across conditions for 
>a given gene) is not a standard use case within edgeR (yet) and 
>there are a number of issues that should be dealt with if you want 
>to do this well.  I'm sure others on the list have more experience 
>and can comment more on this, but it seems you need to get a very 
>good handle on gene length (can you always trust the annotation?) 
>and other biases that (e.g. GC content, positional biases) that may 
>introduce differences in read counts that are different from gene to gene.

Following up on this idea of the problems in comparing expression 
levels between different genes using RNA-Seq data... over a year ago 
I scanned a bunch of articles on RNA-seq and RPKM, and one of them 
actually spiked in a bunch (at least dozens?) of genes at supposedly 
the same concentration. However, when they compared RPKM values, they 
found up to 4 FC differences between these genes with supposedly the 
same concentration! I thought to myself, "Cool!", but as I was 
actually searching for other information, I failed to note which 
paper it was. And now I can't find it again. Anyone know what the 
paper was, or know of other papers that have also tested between-gene 
expression levels? RNA-seq has been touted for it's ability to 
compare expression levels between genes, but as Mark indicated, it 
turns out there are a variety of issues that hinder these comparisons.

Thanks so much!
Jenny




>That said, edgeR has fairly general routines to analyze count data, 
>if you can sort out these other factors.
>
>Cheers,
>Mark
>
>On 2011-01-29, at 10:37 AM, Sridhara Gupta Kunjeti wrote:
>
> > Opps I am sorry there is a typo in the subject - expression
> > Thanks,
> > Sridhara
> >
> >
> >
> > On Fri, Jan 28, 2011 at 6:37 PM, Sridhara Gupta Kunjeti
> > <sridhara at udel.edu>wrote:
> >
> >> Hi All,
> >> I am using RNA-seq to study the expression levels of gene globally and as
> >> well as group of genes. Plotsmear have given me a meaningful 
> information. I
> >> believe the estimated norm.factor is directly included in the 
> model to test
> >> DE, but the values themselves are not modified. My question is 
> if I want to
> >> check compare the expression between genes (within a group) what 
> values do I
> >> need to consider for this analysis. Any help or suggestions will be highly
> >> appreciated.
> >>
> >> Thanks in advance!
> >> Sridhara
> >>
> >>
> >>
> >> --
> >> Sridhara G Kunjeti
> >> PhD Candidate
> >> University of Delaware
> >> Department of Plant and Soil Science
> >> email- sridhara at udel.edu
> >> Ph: 832-566-0011
> >>
> >
> >
> >
> > --
> > Sridhara G Kunjeti
> > PhD Candidate
> > University of Delaware
> > Department of Plant and Soil Science
> > email- sridhara at udel.edu
> > Ph: 832-566-0011
> >
> >       [[alternative HTML version deleted]]
> >
> > _______________________________________________
> > Bioconductor mailing list
> > Bioconductor at r-project.org
> > https://stat.ethz.ch/mailman/listinfo/bioconductor
> > Search the archives: 
> http://news.gmane.org/gmane.science.biology.informatics.conductor
>
>------------------------------
>Mark Robinson, PhD (Melb)
>Epigenetics Laboratory, Garvan
>Bioinformatics Division, WEHI
>e: mrobinson at wehi.edu.au
>e: m.robinson at garvan.org.au
>p: +61 (0)3 9345 2628
>f: +61 (0)3 9347 0852
>------------------------------
>
>
>______________________________________________________________________
>The information in this email is confidential and intend...{{dropped:6}}
>
>_______________________________________________
>Bioconductor mailing list
>Bioconductor at r-project.org
>https://stat.ethz.ch/mailman/listinfo/bioconductor
>Search the archives: 
>http://news.gmane.org/gmane.science.biology.informatics.conductor