[R] covariance analysis by using R

[a.menicacci@fr.fournierpharma.com]

Hi,

Is anyone has solved MR Xin Meng problem (see below) ?

We have the same analysis configuration : 10 groups (including control one)
with 2 mesures for each (ref at t0 and response at t1).

We expect to compare each group response with control response (group 1)
using a multiple comparison procedure (Dunnett test).

In order to perform this test, we have to normalize our data (as you) to
correct response values by the base line normalized.

Covariance analysis seems to represent the best way to do this. But how to
perform this by using R ?

So, if  someone is able to deal with this problem, could you please share
with us your precious knowledge ?

Thanks in advance,

Best Regards.




Alexandre MENICACCI
Bioinformatics - FOURNIER PHARMA
50, rue de Dijon - 21121 Daix - FRANCE
a.menicacci at fr.fournierpharma.com
tél : 03.80.44.76.17



Original message :

Hello sir:
Here's a question on covariance analysis which needs your help. There're 3
experiments,and x refers to control while y refers to experimental result.
The purpose is to compare the "y" values across the 3 experiments.


experiment_1:
x:0.1 0.2 0.3 0.4 0.5
y:0.5 0.6 0.6 0.7 0.9


experiment_2:
x:1 2 3 4 5
y:3 4 6.5 7.5 11


experiment_3:
x:10 20 30 40 50
y:18 35 75 90 98


Apparently,the control("x") isn't at the similar level so that we can't
compare the "y" directly through ANOVA. We must normalize "y" via "x" in
order to eliminate the influence of different level of "x". The method of
normalize I can get is "covariance analysis",since "x" is the covariant of
y.


My question is:
How to perform "covariance analysis" by using R? After this
normalization,we can get the according "normalized y" of every "original
y".


All in all,the "normalized y" of every "original y" is what I want indeed.


Thanks a lot!


My best regards!