Hi. Fold change make sense in paired only, so you need to make it for every pair, like 30min vs 1 h, 30min vs 2h ... for this i would recommend to use some different model, like limma in R. Then you can have linear models for wvery genes and see the tendencies. 

Thank you @Gergo, I have 10 ligands(LPA, PAFF ........) and each ligand has expression w.r.t time (30min, 1hr,2 hr, 4 hr). I calculate the fold change for each ligand (i.e: 10X4=40 columns of fold change), Now I want to generate heat map from it, then it will be ok and right way? OR I need to add something more....?  

A heatmap of fold changes is very descriptive and very nice to look at.

But uf you want to identify whether the changes over time are significant you you need to use statistics. 

It is not clear to me whether you have biological replicates of your experiments? If not doing statistical analysis becomes quite tricky...

If you have biological replicates I would also recommend the tool limma which actually enables you to do an ANOVA style test (take a look at their vignette) meaning you get one p-value pr ligand.

Limma need log2 transformed data as input and assumes normal distribution. This means you should really make sure that you data is normal distributed after log2 transformation. If not you cannot use limma.

Hope this helps

/Kristoffer

Dear Kristoffer,

Thank you for detail answer. I have 2938 differential express gene (DEG) generated through significance analysis of Microarray (SAM). The data is normalized log2 ratio and fold change is calculated and put in excel file. In column we have (control "UNTR" and 10 ligands w.r.t time 30min,1hr, 2 hr and 4 hr) and gene symbols are given in row form. I want to identify the significant ligands and genes from the given file. I am using MeV for analysis and heatmap generation. As an example I attached the excel file. 

Thanks

Dear Muhammad

Do I understand you correctly if the 10 ligands are already significant according to SAM analysis? If that is the case what is it you want to additionally do? 

If your want additional analysis and your data originates from a microarray analysis I would highly recommend that you go back and make the extended analysis you want directly on the microarray data (limma is state-of-the-art for microarray analysis) instead of working on the processed data. This will both allow you to have much more statistical power to detect changes and furthermore the statistical approach will be much more robust (much smaller change you get wrong results and that reviewers will challenge the approach).

Hope this helps

/Kristoffer

Dear Kristoffer,

Thank you very much for your valuable answers. I have 33 ligands in total, which were analysed through SAM. Reported in article entitled "Analysis of the major patterns of B cell gene expression changes in response to short-term stimulation with 33 single ligands".  I selected 10 ligands from above data and wants additional analysis but they didn't provide the RAW data, I downloaded the Processed data from "ArrayExpress". I reviewed the limma tutorial and want to make sure the downloaded data file for limma. I need a starting point for analysis through limma, I attached processed data file as example, Can I use such kind of files as a input for limma and which type of analysis will be performed? 

I really thanks for your answers. 

Dear Muhammad

Unfortunatly I cannot help you with that as I have not done that before. But I think you can get help if you make a new question of how to analyze processed data from arrayexpress.

/Kristoffer