[BioC] fold change, low count reads filter

[Gordon K Smyth]

> Date: Wed, 21 Nov 2012 11:23:55 -0800 (PST)
> From: "Vittoria Roncalli [guest]" <guest at bioconductor.org>
> To: bioconductor at r-project.org, roncalli at hawaii.edu
> Subject: [BioC] 
>
>
> Hi,

> I am new user of edgeR for gene expression analysis. I am testing three 
> different condition, and with pairwise comparison between them I have 
> been able to get the DGE genes for each comparison.

> I have two questions:
>
> 1) I ran the DGE analysis with and without the low count reads filter. 
> When I did not use a filter, I obtained thousand of DGE genes and I 
> could only use a P<0.0001 in order to get a reasonable number to take a 
> look. When I filter the low count reads, as in one of the case studies 
> in the manual, I selected 100cpm in the 3 replicates I have, as cutoff. 
> In this case, with a P value of 0.05 I got few genes from 30-75, 
> depending on the comparison.

It's a good idea to follow the advice in the User's Guide.

> I am a little bit worried that for a global gene expression analysis, 
> this number is too low.

There is no such thing as too many or two many DE genes.  The whole 
purpose of an analysis to find out the truth, whatever that might be.

> Does anyone used a different cutoff for the filtering? Should I try 50 
> cpm?

Without knowing your sequencing depth, how can we tell?  Why not follow 
advice in the User's Guide?

> 2) I am a little bit confused on how the FC is calculated. Is the log2 and the cutoff is FC>2 fold?

Please read the documentation.

> Thanks for the help, I really appreciate your help
>
>
> -- output of sessionInfo():
>
> R version 2.15.1 (2012-06-22)
> Platform: x86_64-apple-darwin9.8.0/x86_64 (64-bit)
>
> locale:
> [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8
>
> attached base packages:
> [1] stats     graphics  grDevices utils     datasets  methods   base

The session information is supposed to be from your edgeR analysis 
session.  Otherwise it gives us no information.

Gordon

______________________________________________________________________
The information in this email is confidential and intend...{{dropped:4}}