This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Deseq2 5 level condition - building contrast

Hello I need your help with my analysis.

So i have an bulk Rna-seq dataset which contains data from ips cells which develop into neurons in 5 days For each time point I have 7 replicates

So now with deseq2 I have made all the pair-wise comparisons and I want to continue by comparing each time point to all the others (e.g time point 1 vs all the other time points) in order to find specific genes for each developmental stage.

I have seen several ways online to do so but I am a bit confused.

Any ideas how t build these kind of contrasts ?

Thanks in advance, A

rna-seq deseq2
• 1,285 views • link
updated by ATpoint ATpoint • written by annaA • annaA

I have seen several ways online to do so but I am a bit confused.

What confuses you, what did you try? Please understand that "it does not work"-like questions are hard to debug. Did you read the DESeq2 manual on contrasts?

• 0 views • link
written by ATpoint ATpoint

So my code until now is as follows

 deseq2Data <- DESeq2::DESeqDataSetFromMatrix(countData=countdata_m, colData=coldata, design= ~ condition)

my reference is 0_dpi

dds <- DESeq2::DESeq(deseq2Data)

pairwise comparisons

res_0vs1 <- results( dds, contrast = c("condition", "0_dpi", "1_dpi") )
res_1vs2 <- results( dds, contrast = c("condition", "1_dpi", "2_dpi") )
res_2vs3 <- results( dds, contrast = c("condition", "2_dpi", "3_dpi") )
res_3vs4 <- results( dds, contrast = c("condition", "3_dpi", "4_dpi") )

After I am filtering the results of each comparison with the following thresholds

padj.cutoff <- 0.01
lfc.cutoff <- 1

Then I tried for example to compare 0_dpi to all otehr time points by this

res <- results(dds, contrast=list("condition_0_dpi", c("condition_1_dpi", "condition_2_dpi","condition_3_dpi","condition_4_dpi"),listVars=c(1, -1/4,-1/4,-1/4,-1/4)))

this didnt work

Also I tried the LRT method with reduced model like

ddsLRT <- DESeq(dds, test="LRT", reduced = ~ 1)
• 0 views • link
written by annaA • annaA

My question is which is the right way to go?

• 0 views • link
written by annaA • annaA

this didnt work

As said, this is not informative. What did not work, why, which errors, what is the problem?

• 0 views • link
written by ATpoint ATpoint

0 answers

No answers yet.

Log in to answer this question.

More posts like this