Question

FEM : How to compute association statistics (limma)

0

Entering edit mode

Nicolas Rosewick ▴ 10

@nicolas-rosewick-10121

Last seen 4.2 years ago

Belgium/Brussels/ULB

Hi,

In the FEM vignette it's written (page 7) : https://www.bioconductor.org/packages/release/bioc/html/FEM.html

" For each gene g in the maximally connected subnetwork, we then derive a statistic of association between its DNA methylation profile and the phenotype of interest (POI) (here normal/cancer status), denoted by t(D,g) as well as between its mRNA expression pro le and the same POI, which we denote by t(R,g). These statistics have already been computed beforehand using the limma package. "

Can someone explain me how to do that using limma ?

So I've two matrix (one containing the methylation state of each gene and samples, one containing the expression of each gene and samples).

Thank you

FEM limma • 1.2k views

ADD COMMENT • link updated 7.9 years ago by andrew • 0 • written 7.9 years ago by Nicolas Rosewick ▴ 10

score 0 · Answer 1 · 2016-06-03

Hi Nicolas, page-7 of the vignette refers to the "in-built" example provided with the package. In practice, for your own data, you would generate the statistics of differential methylation and differential expression using the GenStatM and GenStatR functions, respectively. These functions use limma, but the package is flexible enough for you to compute the statistics in the way you want. Hope this helps. kind regards A -----Original Messages----- From: "nicolas.rosewick [bioc]" <noreply@bioconductor.org> Sent Time: Friday, June 3, 2016 To: andrew@picb.ac.cn Cc: Subject: [bioc] FEM : How to compute association statistics (limma) Activity on a post you are following on support.bioconductor.org User nicolas.rosewick wrote Question: FEM : How to compute association statistics (limma): Hi, In the FEM vignette it's written (page 7) " For each gene g in the maximally connected subnetwork, we then derive a statistic of association between its DNA methylation profile and the phenotype of interest (POI) (here normal/cancer status), denoted by t(D,g) as well as between its mRNA expression pro le and the same POI, which we denote by t(R,g). These statistics have already been computed beforehand using the limma package. " Can someone explain me how to do that using limma ? So I've two matrix (one containing the methylation state of each gene and samples, one containing the expression of each gene and samples). Thank you Post tags: FEM, limma You may reply via email or visit FEM : How to compute association statistics (limma)