Search code examples
rdrmdrc

R, drc package's drm() function display negative hillslope. even though the graph describe to increasing feature

I used to analysis the results by GraphPad Prism. these days I try to analysis by using R & drc packages.

When I try drm(...., fct.LL4()) I find the hill slope (b factor) is negative value even though the graph is increasing.

Surprisingly the GraphPad Prism show me the positive hill slope against same graph with drc.

Some people said to drm's hill slope is kinds of relative value, so just catch the maning of value.

But in biological field, hill slope is very important factor because it imply that drug's operation.

So, can I get advice why drc package is designed to express negative hill slope?

Solution

  • The drc package uses the following parameterization of the four-parameters logistic curve:

    enter image description here

    This curve is decreasing when b > 0:

    f <- function(x, b, c, d, e){
  c + (d-c)/(1+exp(b*(log(x)-log(e))))
}
x <- seq(1, 10, by = 0.1)
y <- f(x, b=20, c=0, d=2, e=5)
plot(x, y, type = "l")

    enter image description here

    If you are used to have a positive b when the curve is increasing, that's because of a different parameterization.