[BioC] Linear response gene selection techniques

[James W. MacDonald]

david neil hayes wrote:
> Is there a technique in Bioconductor for selecting genes associated with 
> a linear response such as chemotherapy response (IC50), while accounting 
> for false discovery rates/multiple testing?  I could not find an 
> explicit procedure with this intent.

I believe you can use the limma package for this sort of analysis. The 
paradigm is to use lmFit() to fit an ANOVA, but as this is just a 
wrapper to lm.fit(), I don't see any reason offhand that would preclude 
using lmFit() for a continuous response.

Since the paradigm is to use lmFit() for ANOVA, most of the 
documentation has the following code:

design <- model.matrix(~ -1 + factor(c(1,1,1,2,2,2)))

Which will fit an ANOVA with two factor levels and no intercept. To fit 
a continuous variable you would remove the factor() statement and 
instead use the amounts of chemo drug you used. I would also recommend 
fitting an intercept unless you have a reason to believe the regression 
line should go through (0, 0).

design <- model.matrix(~ c(0.1, 0.5, 1.0, 1.5, 2.0))

where 0.1, ... 2.0 are the concentrations of your drug.

Best,

Jim


> 
> Neil Hayes
> 
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor


-- 
James W. MacDonald
Affymetrix and cDNA Microarray Core
University of Michigan Cancer Center
1500 E. Medical Center Drive
7410 CCGC
Ann Arbor MI 48109
734-647-5623