Problem in DESe2 with multiple factors

usegalaxy.eu support
, ,
1

Hi all,

I am currently analysing my data from an RNA-seq experiment. I have brain tissue for mice, and I aim to find differentially expressed genes. I am using the DESeq2 tool, and I follow the instructions of the tutorial: Reference-based RNA-Seq data analysis ( Analysis of the differential gene expression)

My experimental design includes the right and left cortex, five time points (control, 1d, 3d, 5d, 10d) and both sexes.
Initially, I split the right and the left cortex. In this case, I had five time points as the main factor and sex as a secondary factor, which worked perfectly.

But then, I wanted to combine the cortex with the time points and leave the sex as the second factor. In this case, I will have something like 10d right, 10d left, 5d right, 5d left, etc. ( I hope the picture is visible, and it shows what I did)

Screenshot_problem
Screenshot_problem2560×932 276 KB

However, this gives me the error:

Error in data.frame(sample = basename(filenames_in), filename = filenames_in, :
duplicate row.names: /data/dnb09/galaxy_db/files/c/a/0/dataset_ca0e0782-ed41-46b8-82bd-dde5b1d760a9.dat, /data/dnb09/galaxy_db/files/7/1/b/dataset_71b67376-3eb5-4d35-99b7-ea7b055190aa.dat, /data/dnb09/galaxy_db/files/5/8/e/dataset_58e20c2b-5d72-4c12-a7a2-62821d206776.dat, /data/dnb09/galaxy_db/files/3/9/1/dataset_3911a4cb-d1bc-4c9b-9067-77616683ff97.dat, /data/dnb09/galaxy_db/files/1/b/e/dataset_1be6bb3e-cc36-4061-8366-12ac28f1b81a.dat, /data/dnb09/galaxy_db/files/8/2/e/dataset_82e3626f-6786-458b-b295-66714fc1065d.dat, /data/dnb09/galaxy_db/files/d/b/e/dataset_dbed6aa2-07c6-4a46-bb43-1b9927095de3.dat, /data/dnb09/galaxy_db/files/0/7/b/dataset_07bdaf17-e378-4752-9303-6ad77bbca803.dat, /data/dnb09/galaxy_db/files/e/a/3/dataset_ea363802-45fa-4e7e-8deb-c4c9b41fba5c.dat, /data/dnb09/galaxy_db/files/1/7/1/dataset_171dc26c-1ba7-4322-81be-33cb05137511.dat, /data/dnb09/galaxy_db/files/d/5/1/dataset_d513cb1d-721f-49be-b4dc-d9927328197e.dat, /data/dnb09/galaxy_db/files/4/f/1/dataset_4f10

I have enough biological replicas per condition, so this should not be a problem,

I have participated in some courses and watched some videos/tutorials regarding the analysis in Galaxy, and in principle,it is possible to combine multiple tags, but I never managed to do it!

Should I change any parameters? Is there something wrong with my approach?

I appreciate any comments or ideas!! :slight_smile:

2

The tags let you select those datasets in a collection that are tagged. So you can’t select the same tag in a contrast for the same factor (i.e I don’t think you can select 10d more than once). You can however add multiple factors, where the first one is your primary factor, the next is the secondary factor, the third is the tertiary factor and so on.

3

Thank you for you answer!! :slight_smile:

The problem is that the cortex (right and left)is not a secondary factor. My research question in the end would be something like: Which are the differential expressed genes between the time points (control, 1d, 3d, 5d, 10d)and simultaneously between the cortex of the same time for example compare 10 d right vs 10d left?

Do you think that this is possible in Galaxy?

4

You’ll have to look at the deseq2 vignette at Analyzing RNA-seq data with DESeq2 to find out how to do this for time series data.

I don’t know if I can translate your question into a formula, but you can always pick out the contrasts one by one.

The problem is that the cortex (right and left)is not a secondary factor.

It does sound like a secondary factor if your primary question is which genes are differentially expressed between timepoint A and timepoint B.