Re: Paired limma analysis

Please don't take things off-list.

On 8/3/2012 12:21 PM, mb3058@... wrote:
> Hi Jim,
>
> Thank you for replying to my questions. Your answers have somewhat 
> cleared the doubts that I had about the results. I will be obliged if 
> you could reply to some follow up questions I had.
>
> 1.Out of Pair2 and Treatlo, which one is the logFC column (my input 
> data is log2 transformed)?

I have no idea what you are asking here.

> 2.What threshold should I be taking for Treatlo coefficient to get 
> genes that are significantly differentially expressed between hi and lo.

That's up to you, isn't it? If you are doing the analysis, then you have 
to take responsibility for the choices you make. Why would you abdicate 
this responsibility to some random dude on a listserv? If I said you 
need a p-value of 0.5 and a -1.5 fold change, would you use that advice? 
I have to admit, I am somewhat boggled.

Best,

Jim

>
> Thanks again
>