Hi all
I need advice on how to analyse my complicated design RNAseq data with Deseq2.
I have 3 biological replicates, 2 time points and a control just for the first time point . That is:
control condition timepoint 1 (x3 biol replicates)
test condition time point 1 (x3 biol replicates)
test condition time point 2 (x3 biol replicates)
I am interested to find genes that increase expression in response to treatment and remain more highly expressed at time point 2 in comparison to the control.The aim is to look for overrepresented features in this gene set. So, keeping low false positive rate is not hugely critical.
I understand that this experimental design is not ideal as we cannot account for technical and other unrelated variation on the second time point. But the data were generated without my advice and now I have to deal with it.
So far I have done 2 pairwise comparisons. 3 vs 3 samples from time point 1 and 3 samples from test condition from time point 2 vs the 3 control samples from time point 1. And then looked for genes that appear upregulated in these 2 pairwise comparisons.
The issue is that expression changes are much smaller in the second timepoint (that was expected from the start) so I am now looking for genes that are significantly upregulated in time point 1 and show even small upregulation (in comparison to control) at time point 2.
Is there a more sensitive way to analyse this data??
I need advice on how to analyse my complicated design RNAseq data with Deseq2.
I have 3 biological replicates, 2 time points and a control just for the first time point . That is:
control condition timepoint 1 (x3 biol replicates)
test condition time point 1 (x3 biol replicates)
test condition time point 2 (x3 biol replicates)
I am interested to find genes that increase expression in response to treatment and remain more highly expressed at time point 2 in comparison to the control.The aim is to look for overrepresented features in this gene set. So, keeping low false positive rate is not hugely critical.
I understand that this experimental design is not ideal as we cannot account for technical and other unrelated variation on the second time point. But the data were generated without my advice and now I have to deal with it.
So far I have done 2 pairwise comparisons. 3 vs 3 samples from time point 1 and 3 samples from test condition from time point 2 vs the 3 control samples from time point 1. And then looked for genes that appear upregulated in these 2 pairwise comparisons.
The issue is that expression changes are much smaller in the second timepoint (that was expected from the start) so I am now looking for genes that are significantly upregulated in time point 1 and show even small upregulation (in comparison to control) at time point 2.
Is there a more sensitive way to analyse this data??