Contrast and targets.txt

[ahmerjamil]

First of all thanks a lot for this nice and easy to use pipeline. I am using it to analyze my samples but I would like to ask you that for DEAnalysis you mentioned two files targets.txt and contrast.txt but I am not sure how it should be look like. Could you please give me an example contrast.txt and targets.txt file.

For example samples are like following

Cancer_1, Cancer_2, Cancer_3 Healthy_1,Healthy_2,Healthy_3

I would like to see the differential expression between cancer and healthy.

Thakns in advance.

Best Regards, Muhammad Jamil

[Eduardo Andres Leon]

Dear Muhammad Jamil, In the examples webpage (http://miarmaseq.cbbio.es/examples), you have several examples (for miRNA, mRNA and circRNA).

I paste you the miRNA example:

targets.txt (each column is tab separated):

Filename        Name    Time
SRR873382       Norm_1  T0h
SRR873383       Norm_2  T0h
SRR873384       Hyp_16_1        T16h
SRR873385       Hyp_16_2        T16h
SRR873386       Hyp_32_1        T32h
SRR873387       Hyp_32_2        T32h
SRR873388       Hyp_48_1        T48h
SRR873389       Hyp_48_2        T48h

contrast.txt

 "Name"
 "Comp_16=T16h-T0h"
 "Comp_32=T32h-T0h"
 "Comp_48=T48h-T0h"

As you see, the contrast.txt contains three different comparisons.

Let me know

[ahmerjamil]

Thanks a lot. So it means I can have multiple comparison within one run?

[Eduardo Andres Leon]

yes ;)

[ahmerjamil]

Can we have contrast.txt as follow also?

"Time" "Comp_16=T16h-T0h" "Comp_32=T32h-T0h" "Comp_48=T48h-T0h" "Name" "Norm=Norm_1-Norm_2"

[Eduardo Andres Leon]

Ummm I can't see very well your file, but I guess that you mean to make two different comparison from 2 different columns. If that it is true, I'm not really sure (I've never test it). Take one thing into account. Once you have successuflly finished one section (eg: Quality, ReadCount, TargetPrediction ... whatever), you don't need to do it again. That means that, if you have aligned for files, you can comment the whole section in the ini file and this process will not be executed. So, what I mean is that, once you have your bam files and your tab files (ReadCount), you can do as many DEAnalysis (~3 minutes) as you want.

[ahmerjamil]

Its not working the way I was doing it. But I found that for contrast only last column is considered so I can add as many analysis in last column.

Also I recommend to add a batch correction method using svaseq(sva package bioconductor) in pipeline as my comparison giving results which is due to batch effect rather than actual differences.

Hope you add it soon as I am doing it manually at the moment.