Entering edit mode
Suresh, Uthra
▴
10
@suresh-uthra-4009
Last seen 9.6 years ago
Hello List,
I am currently working on Affy arrays for a time course experiment. I
have two groups (Control and Treated) with samples taken at 1hr, 24hrs
and 48hrs.
The following is how my targets file looks like:
FileName Targets
Files 1-4 control.0hr
Files 5-8 control.1hr
Files 9-12 treatment.1hr
Files 13-16 control.24hr
Files 17-20 treatment.24hr
Files 21-24 control.48hr
Files 25-28 treatment.48hr
I followed the limma user guide (pg48 and 49) to set up the design
matrix for the lmFit test. A part of the design matrix is given
below.
Design:
fcon.0hr fcon.1hr fmms.1hr fcon.24hr fmms.24hr fcon.48hr fmms.48hr
1 1 0 0 0 0 0 0
2 1 0 0 0 0 0 0
3 1 0 0 0 0 0 0
4 1 0 0 0 0 0 0
5 0 1 0 0 0 0 0
6 0 1 0 0 0 0 0
7 0 1 0 0 0 0 0
8 0 1 0 0 0 0 0
I have 7 columns and 28 rows in my design matrix.
I am now interested to look at the differentially expressed genes
between the control and treated samples across different time points.
I have the following questions:
1. Should my contrast matrix be,
cont.dif<-makeContrasts(Dif1hr=(fmms.1hr-fcon.0hr)-(fcon.1hr-
fcon.0hr),
Dif24hr=(fmms.24hr-fmms.1hr)-(fcon.24hr-
fcon.1hr),
Dif48hr=(fmms.48hr-fmms.24hr)-(fcon.48hr-
fcon.24hr),
levels=design)?
2. If yes, does Dif24hr give those genes that are differentially
expressed between the treated and control samples from time 1hr to
24hrs and similarly for Dif48hr?
3. Can the following contrast matrix also be used?
cont.dif1<-makeContrasts(Dif1hr=(fmms.1hr- fcon.1hr),
Dif24hr=(fmms.24hr-fcon.24hr),
Dif48hr=(fmms.48hr- fcon.48hr),
levels=design)? (this contrast matrix completely
ignores the files for the control samples at time 0hr? can this be
correct?)
I would really appreciate if someone can suggest an appropriate method
to deal with this kind of experimental design.
Thank you,
Regards,
Uthra Suresh
UTHSCSA
[[alternative HTML version deleted]]