Hi,
I am just starting working with DESeq. I have a question regarding the basic biological interpretation of DESeq based DE gene expression. There are two situations I have listed below and I would like to know which one is more biologically relevant
I have two treatment groups: treatment 1 and treatment2 and I am comparing them with a control group all with three replicates. I devised my study as
1. I created a dataframe containing counts of all 9 count files and from this dataframe, I am creating comparisons as: T1 vs Control, T2 vs Control and T2 vs T1.
2. I create a dataframe everytime I create a comparison like when I am comparing T1 vs Control, then I am creating a dataframe with 6 count files. Again when I am comparing T2 vs Control I am creating another dataframe with 6 count files.
I want to know which of these two design strategies will give me a more accurate result as to what effects T1 and T2 are causing when compared with control and how are T1 and T2 different as well as similar?
Hi Mrigaya,
I would certainly not split the analysis in multiple data frames. DESeq will estimate some negative binomial parameters based on your data, and the more data you provide, the more reliable the estimates. Also, you want to work with the same estimates for all comparisons.
There are ways of putting the 3 comparisons into one model but the specific details depend on which contrasts you are interested in. I wouldn't know the exact syntax, but I'm sure that the community can help you if you specify precisely what you want to test.
By the way: there are many good software packages for the detection of differential expression. For more complex designs, I personally really like EBseq, which has a very straightforward syntax to test multiple hypothesis. Just saying.
Good luck!
Rik
Hi Mrigaya,
I would certainly not split the analysis in multiple data frames. DESeq will estimate some negative binomial parameters based on your data, and the more data you provide, the more reliable the estimates. Also, you want to work with the same estimates for all comparisons.
There are ways of putting the 3 comparisons into one model but the specific details depend on which contrasts you are interested in. I wouldn't know the exact syntax, but I'm sure that the community can help you if you specify precisely what you want to test.
By the way: there are many good software packages for the detection of differential expression. For more complex designs, I personally really like EBseq, which has a very straightforward syntax to test multiple hypothesis. Just saying.
Good luck!
Rik
Rik is correct. Dispersion estimates will be much more reliable if you include all nine samples in the analysis.
Hi Mrigaya,
as Rik has explained the best option is to include in a single analysis all your comparisons, to reduce the data dispersion and estimation of DE genes.
There are many softwares that allows you do to it, like the EBSeq, RobiNA, Galaxy, ... among others. Which by de way could allow you to use also edgeR to determine DE genes, as in some ones DEseq performs the comparison in inverse order control vs test, and you have all your data with opposite sign (meaning, up-regulated are shown as down-regulated and also at the inverse, down-regulates as up-regulated) while the shorting up of DE genes remains constant.
good luck with your analysis,
Beatriz
- Badges
- Science topic
- Similar topics
- Biological Science
- Genetics
- Gene Expression