This question has multiple aspects:
1) I would like to know how to put a P value on an EC50 shift: "e.g. I have a bradykinin dose response curve in my control group and now I have block eNOS using LNAME so the curve shifts." Two-way ANOVA would give me the difference between both curves at all time points indiscriminately.
2) I repeat this experiment in my second and third groups and I want to compare the effect of LNAME versus bradykinin within each group to the other groups (assuming the bradykinin response itself is also altered).
Hello Vincent,
The usual way to design and evaluate an experiment of this type is with a generalized linear model (references provided below). The normal assumption is that the dose responses collected are independent and approximately Logistic (Probit is roughly equivalent) in distribution,
A step by step explanation of how to do this using free software from the R Language for Statistical Computing (http://www.r-project.org) can be found on pages 190 - 199 of "Modern Applied Statistics with S" by Venables and Ripley (2002). [This may not be an easy read for you. It wasn't an easy read for me when I first saw it several years ago.]
Useful information on how to select doses during the design of a dose response study can be found on pages 70 - 81 of "Statistical Techniques in Bioassay" by Z. Govindarajulu (2001).
The classic reference for this problem is "Statistical Method in Biological Assay" by Finney (1978).
Is this helpful or am I missing some aspect of your question?
Chuck
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Applied Statistics