[R] DRC: Effective doses versus Predicted values

[Iban]

Hi!
I want to use the DRC package in order to calculate the IC50 value of an
enzyme inhibition assay.

The problem is that the estimated ED50, is always out of the fitted curve.

In the example below, I had a ED50 value of 2.2896, 
But when I predict the response level for this concentration I get a value
of 45.71 instead of the expected value of 50.

This is my data:

#Dose unit is concentration (mM); response  unit is % of inhibition

dat <- data.frame(dose=c(0.480,0.957,2.870,4.780,6.700,9.570), 
response=c(11.88171,28.62986,51.50847,65.51793,77.29431,82.83489))

model <- drm(response ~ dose, data=dat, fct= W2.3())

ED(model,50)
#get a ED50 valu of2.2896

predict(model, data.frame(dose=2.2896))
#get a response level of  45.714

plot(model)
points(2.2896,50, col="red")
#the ED50 value is out of the fitted curve
-- 
View this message in context: http://r.789695.n4.nabble.com/DRC-Effective-doses-versus-Predicted-values-tp2322613p2322613.html
Sent from the R help mailing list archive at Nabble.com.