From: John S. Walker <jsw9c_at_uic.edu>

Date: Fri 16 Jun 2006 - 01:36:39 EST

R-help@stat.math.ethz.ch mailing list

https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html Received on Fri Jun 16 01:40:53 2006

Date: Fri 16 Jun 2006 - 01:36:39 EST

Gday,

The dataframe is structured like this:

expt treatA treatB dose force. 1 - - 0.1 20 1 - - 0.2 40 ... 4 + + 0.1 20 4 +

I used a groupedData object: mydata=groupedData(force ~ dose | expt)

I used an nlme obect to model the data as follows (pseudocode):

myfit.nlme <- nlme(force ~ ss_tpl(dose, upper, ed50,slope), fixed=list(ed50~factor(treatA)*factor(treatB)))

The function ss_tpl is a properly debugged and fully functional
selfstarting three parameter logistic function that I wrote- no problem
here. In my analysis

I also included fixed terms for the other fit parameters; upper and
slope, but my main problem is with the
ed50 so that's all I've included here.

Running an anova on the resulting object (anova(myfit.nlme) I found the
A -/B- (control) to

be significantly different from zero, treatment A was significantly
different, treatment B had no significant
effect and there was a significant interaction between treatment A and
treatment B.

The interaction term is likely to be real. The treatments are on sequential steps in a pathway and treatment B may be blocking the effect of treatment A, i.e. treatment B alone has no effect because it blocks a pathway that is not active, treatment A reduces force via this pathway and treament B therefore blocks the effect of treatment A when used together.

From what I understand, please correct me if I'm wrong, the parameter estimates from summary(model.nlme) are not correct for main effects if a significant interaction is present. For example in my data treatment B alone has no signifcant effect in the anova but the interaction term A:B is significant. I believe The summary estimate for B is the estimate across all levels of A. What I want to do is pull out the estimate for B when A is not present. I suppose I can do it manually from the list of coefficients from nls or fit a oneway model with treatment levels A, B, AB. But I was kind of hoping there was some extractor function.

The reason I need this is that the co-authors want to include a table of parameter values with std errs or confidence intervals ala:

Treat upper ed50 slope A-/B- x x x <- shows value for comparison to control studies A+/B- x x x <-Shows A is working0 A-/B+ x x x <- Shows B has no effect alone A+/B + x x x <-shows B blocks A (not necessarily total)

So back to my question,How do I extract estimates of the parameters
from my model object for a

specific combination of factors including the interaction term.

i.e. what is the ed50 (and std err) for A-/B-, A+/B-, A-/B+, A+/B+ ?

I think this is a fair question and one that many biomedical scientists would need.

R-help@stat.math.ethz.ch mailing list

https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html Received on Fri Jun 16 01:40:53 2006

Archive maintained by Robert King, hosted by
the discipline of
statistics at the
University of Newcastle,
Australia.

Archive generated by hypermail 2.1.8, at Fri 16 Jun 2006 - 04:12:57 EST.

*
Mailing list information is available at https://stat.ethz.ch/mailman/listinfo/r-help.
Please read the posting
guide before posting to the list.
*