Question: understanding DEseq2 interactions before/after releveling the baseline
gravatar for assaf www
14 days ago by
assaf www140
assaf www140 wrote:

Possibly this is a trivial question, I'm missing something:

My experiment is based on the model

~ pos +  ring + pos:ring

the factors can take the following values:

pos = {Base,Junction,Tip} 
ring = {Down,Up,Top}

I made two DESeq2 runs, where in the first run I used the default levels of factors (as above),and in the second run I releveled the order of ring to {Up,Down,Top}. When testing the effect of "pos_Junction_vs_Base", I got 7 and 915 signifcant genes in run1 and run2, respectively. So, if my interpretation is correct then at ring=Down only 7 genes differ between Junction and Base, compared to 915 at ring=Up. If the above is correct then I expect that the interactions "posJunction.ringDown" and "posJunction.ringUp" will give about 900 signifcant genes each, to reflect the difference in Junction vs Base response between ring=Down and ring=Up. But I see only 3 signifcant genes for these interactions.

thanks a lot for any help !

deseq2 • 97 views
ADD COMMENTlink modified 14 days ago by Michael Love24k • written 14 days ago by assaf www140

Just make it easier for people to help you. Write out explicitly how everything is leveled at each step, every design command line, and every results call.

ADD REPLYlink written 14 days ago by swbarnes2240
Answer: understanding DEseq2 interactions before/after releveling the baseline
gravatar for Michael Love
14 days ago by
Michael Love24k
United States
Michael Love24k wrote:

There is an important difference between significant for baseline comparison X, significant for baseline comparison Y, and then testing the interaction.

One way to think about this is, suppose that the "Down" samples have a lot of noise, or there are very few of them, relative to the "Up" samples. So you have more precision to estimate the position effect for the Up samples (more significant DEGs), but less precision for estimate the effect in the Down samples. When you go to estimate the interaction, you actually need precision in both groups, in order to determine if the interaction is significant. See for example, the first two lines here:

ADD COMMENTlink written 14 days ago by Michael Love24k

makes sense, thanks a lot !!!

small related question: in addition to lower power of detecting interaction effects than main effects (as I think the introduction of this article says), can such results reflect, for example (among other options), that samples in Downbase tend to have larger biological variation than samples in Upbase, which in turn lead to lower detection power of Downbase vs. Downjunction changes (than Upbase vs. Upjunction changes) ? I can try estimating such options visually with NMDS, can I also test this from DEseq2 output at group/gene level for example, or by other related DEseq2 tests ?

ADD REPLYlink modified 13 days ago • written 13 days ago by assaf www140

Yes, that's what I was trying to get at with my first reply above. There are many reasons why you may not have a significant interaction tests despite having significant baseline differences. Hope this answers the original question about why you see differences in your various results tables.

ADD REPLYlink written 13 days ago by Michael Love24k
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 16.09
Traffic: 280 users visited in the last hour