If you are not familiar with R that will be difficult...

The limma-function you will need is lmFit. It requires the data as log expression values in a matrix, with genes in rows and samples in columns. A further argument passed to lmFit is the design of your study (i.e. the linear model). You can read the help and the vignette to limma to find out how you can specify the design matrix.

After having fit the model, the limma function eBayes is used to calculate the gene-wise tests statistics (moderated t-stats, p-values and B-stats).

This info won't be enough for you to solve the puzzle. It will take quite a bit more work. You will have to learn a bit of R and understand a bit how limma works. That will take time...

Although I strongly recommend learning R (it is good anyway!) it might really be a too effort just to solve one problem. So here it might be better to ask someone who knows R (and ideally limma, too), to help you and do the analysis for you. For someone who knows these things it should not take longer than 2-5 minutes.

If you are not familiar with R that will be difficult...

The limma-function you will need is lmFit. It requires the data as log expression values in a matrix, with genes in rows and samples in columns. A further argument passed to lmFit is the design of your study (i.e. the linear model). You can read the help and the vignette to limma to find out how you can specify the design matrix.

After having fit the model, the limma function eBayes is used to calculate the gene-wise tests statistics (moderated t-stats, p-values and B-stats).

This info won't be enough for you to solve the puzzle. It will take quite a bit more work. You will have to learn a bit of R and understand a bit how limma works. That will take time...

Although I strongly recommend learning R (it is good anyway!) it might really be a too effort just to solve one problem. So here it might be better to ask someone who knows R (and ideally limma, too), to help you and do the analysis for you. For someone who knows these things it should not take longer than 2-5 minutes.

Jochen – I laughed when I saw your estimate of 2–5 minutes 'for someone who knows these things'

The trouble is the years of work required to be able to do things like this in 2–5 minutes, isnt it?

You are right, for sure :)

But after a few years of training, well, it really is a matter of a few minutes. That's nice about R: once you can handle it (what takes many years) you are able to save many more years of your lifetime solving problems in minutes that would otherwise take days and months ;)

Do you have one control group and one experimental group? If you have multiple control and experimental groups, it might be a better idea to use some type of ANOVA. You should be able to use a 2-way ANOVA and "block" on experiment. Suppose you have 3 control groups (C1, C2, C3) and 3 experimental groups (E1, E2, E3). 

They you will have 3 columns of data:

Experiment Group Result

1                     Cont    xxxx

1                      Exp     xxxx

2                      Cont  xxxx

2                      Exp    xxxx

3                     Cont    xxxx

3                      Exp     xxxx

After looking at your data like this, you should be able to get a steady baseline for comparisons. 

You can even try a "paired" design with your data and use ANOVA. You'll have 2 columns of data. Experiment and "difference" . (C11 - E11), (C12 - E12), etc. 

Experiment   Difference

1                       (C11-E11)

1                       (C12 - E12)

.

.

.

.

3                         (C33 - E33)