Question: challenging design matrix
0
gravatar for bpshiel
2.9 years ago by
bpshiel0
bpshiel0 wrote:

Hi all,

Im using limma for differential expression.

I've been struggling with the concept of my experimental design analysis for a few days so i need some help.

 

My groups are something like this...

Feed condition survival
feed1 wet 10
feed1 wet 4
feed1 wet 6
feed1 wet 4
feed1 dry 3
feed1 dry 3
feed1 dry 4
feed1 dry 5
feed2 wet 10
feed2 wet 8
feed2 wet 5
feed2 wet 7
feed2 dry 9
feed2 dry 5
feed2 dry 6
feed2 dry 8
feed1+ dry 9
feed1+ dry 5
feed1+ dry 8
feed1+ dry 7

For example, this study is aimed at determining the genes that are responsible for higher Survival rates in dry environments when fed feed2 and feed1+ in comparison to  feed1, using a few different comparison between these 5 treatments.

the easiest method to do this i thought would be to just assign the 5 treatments to a group and then compare the comparisons im interested in.

e.g.

Feed1wet, feed1dry, feed2wet, feed2dry and feed1+dry

My real issue, is that i have been struggling with wether to take Survival into account as a continuous variable.

I thought this made sense at first, but now considering my project aim of determining the genes that may be responsible for survival. I think that i will be masking the DE genes I'm interest in.

e.g. if i want to compare DE genes between feed1dry and feed1+dry to understand why feed1+ has higher survival in dry conditions, accounting for suvival as a continuous variable will mask what im interested in won't it?

what do you think?

 

thanks

ADD COMMENTlink modified 2.9 years ago by Gordon Smyth37k • written 2.9 years ago by bpshiel0
Answer: challenging design matrix
1
gravatar for Gordon Smyth
2.9 years ago by
Gordon Smyth37k
Walter and Eliza Hall Institute of Medical Research, Melbourne, Australia
Gordon Smyth37k wrote:

Yes it would.

You could correlate genes directly with survival, but better to do that as a separate analysis without Feed and Condition.

ADD COMMENTlink modified 2.9 years ago • written 2.9 years ago by Gordon Smyth37k
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 16.09
Traffic: 249 users visited in the last hour