Hi,
I have an experiment where I have two different siRNAs (a and b) that
target the same gene. I have 18 two colour microarray experiments
where
the control on Cy3 is siRNA non-targetting. For each siRNA there is
three time points and for each of those time points three replicates
(biological). I am planning on ignoring the time point for now. The
comparisons I would like to make are siRNA a vs control, siRNA b vs
control, both siRNA vs control and a vs b.
So this is the design matrix I have got set up:
a b siRNA
1a_48 1 0 1
1a_72 1 0 1
1a_96 1 0 1
1b_48 0 1 1
1b_72 0 1 1
1b_96 0 1 1
2a_48 1 0 1
2a_72 1 0 1
2a_96 1 0 1
2b_48 0 1 1
2b_72 0 1 1
2b_96 0 1 1
3a_48 1 0 1
3a_72 1 0 1
3a_96 1 0 1
3b_48 0 1 1
3b_72 0 1 1
3b_96 0 1 1
I think I must be doing something wrong as when I run the lmFit()
function I get the following:
> fit <- lmFit(MA, design)
Coefficients not estimable: siRNA
Warning message:
In lmFit(MA, design) :
Some coefficients not estimable: coefficient interpretation may
vary.
Is my design matrix incorrect?
Thanks
--
**************************************************************
Daniel Brewer, Ph.D.
Institute of Cancer Research
Molecular Carcinogenesis
Email: daniel.brewer at icr.ac.uk
**************************************************************
The Institute of Cancer Research: Royal Cancer Hospital, a charitable
Company Limited by Guarantee, Registered in England under Company No.
534147 with its Registered Office at 123 Old Brompton Road, London SW7
3RP.
This e-mail message is confidential and for use by the
a...{{dropped:2}}
I have an experiment design involving 2 varieties with 3 treatments
each (6 sample types). The 3 treatments were compared within each
variety with a direct design for a total of 3 treatments x 2 dye swaps
= 6 for a single variety and 12 for both varieties. We also compared
the 3 treatments across the varieties with a dye swap so 6 (3 x 2)
more comparisons. Each of the 6 sample types was therefore measured 6
times (3 with each dye).
I'm setting up the design matrix based on the direct design example in
the limma guide so we can use more than just the direct comparisons to
estimate the coefficients of interest, which are basically all of the
direct comparisons. I created one "basic" design matrix with just the
direct comparisons to get an idea of what the coefficients should be.
Then I tried to do a design matrix to pull in the other indirect
comparisons for the "within variety" comparisons but the results are
sometimes far from the "basic" design results.
My arrays are:
N R-Cy5 G-Cy3
1 v1.t1 v1.t2
2 v1.t2 v1.t1
3 v1.t2 v1.t3
4 v1.t3 v1.t2
5 v1.t3 v1.t1
6 v1.t1 v1.t3
7 v2.t1 v2.t2
8 v2.t2 v2.t1
9 v2.t2 v2.t3
10 v2.t3 v2.t2
11 v2.t3 v2.t1
12 v2.t1 v2.t3
13 v1.t1 v2.t1
14 v2.t1 v1.t1
15 v1.t2 v2.t2
16 v2.t2 v1.t2
17 v1.t3 v2.t3
18 v2.t3 v1.t3
Following the limma guide for the direct design comparisons (within
varieties - using the t2 sample as the variety "reference") is in the
first 4 columns. The 3 comparisons between varieties are the last 3
columns.
v1t1.v1t2 v1t3.v1t2 v2t1.v2t2 v2t3.v2t2 v1t1.v2t1 v1t2.v2t2
v1t3.v2t3
[1,] 1 0 0 0 0 0
0
[2,] -1 0 0 0 0 0
0
[3,] 0 -1 0 0 0 0
0
[4,] 0 1 0 0 0 0
0
[5,] -1 1 0 0 0 0
0
[6,] 1 -1 0 0 0 0
0
[7,] 0 0 1 0 0 0
0
[8,] 0 0 -1 0 0 0
0
[9,] 0 0 0 -1 0 0
0
[10,] 0 0 0 1 0 0
0
[11,] 0 0 -1 1 0 0
0
[12,] 0 0 1 -1 0 0
0
[13,] 0 0 0 0 1 0
0
[14,] 0 0 0 0 -1 0
0
[15,] 0 0 0 0 0 1
0
[16,] 0 0 0 0 0 -1
0
[17,] 0 0 0 0 0 0
1
[18,] 0 0 0 0 0 0
-1
Is there a reason the coefficients generated by this design are so
different in some cases from the "basic" design?
Thanks for any help,
John Fernandes
[[alternative HTML version deleted]]
