Limma design matrix for time course experiment
0
0
Entering edit mode
@gordon-smyth
Last seen 47 minutes ago
WEHI, Melbourne, Australia

Moderator: This post is an answer to: Limma design matrix for time course experiment

Dear Uthra Suresh,

There is no single shrink-wrapped analysis which is standard and correct for a time course experiment, or for any experiment really.  Rather the appropriate analysis depends on what questions are of interest to you and relevant to the biology of the experiment.  Each analysis is the answer to a particular question.  The section of the Limma User's Guide that you refer to tries to emphasise this point by posing each contrast matrix in terms of a question.

Both of the contrast matrices that you give are corrrect, they just answer
different questions.

The second contrast matrix (cont.dif1) is correct if you want to find genes which are differentially expressed between the treatment and control at each time point.

The first contrast matrix (cont.dif) is correct if you want to find genes which change over time differently between the treatment and the control.

With the second design matrix you are looking for straight differences in expression.  With the first design you are looking for differences between time-specific responses, i.e., differences between changes in expression. (Statisticians call this an interaction effect.) Results from the
second matrix will include genes which are different between control and treatment regardless of time.  The first matrix instaed focuses on time-course changes, mostly subtracting out any baseline differences.

What question do you want to ask?  Only you can decide.

Best wishes
Gordon

affy limma • 1.9k views
ADD COMMENT

Login before adding your answer.

Traffic: 761 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6