CODEWITHSUNDEEP
rggplot2
Sam_9090
Sam_9090

Reputation: 321

Plot drc using ggplot

I would like to plot the dose response curve using ggplot but when I do so it looks very different. Is it possible to plot the curve with ggplot as it looks exactly when it is polotted using plot function from base R?

My second question about the drm model summary: Which parameter is more reflective of the binding: b, c, d, or e?

library(drm)
library(ggplot2)

test_data <- data.frame(Conc = c(0.0004882812, 0.001953125 ,0.0078125, 0.03125 ,0.125 ,0.5),
                        Response = c(1.616017 ,1.165835, 0.5783709, 0.3440007, 0.2668585, 0.2336709))


## plot drm model
model_drm <- drm(Response ~ Conc, data=test_data,
                 fct=LL.4())
summary(model_drm)
plot(model_drm)


## now using ggplot

  ggplot(test_data, aes(x = Conc, y = Response )) +
  geom_point() +
  stat_smooth(method = "drm",
              method.args = list(
                fct = LL.4()),se = FALSE)

enter image description here

enter image description here

Model fitted: Log-logistic (ED50 as parameter) (4 parms)

Parameter estimates:

                Estimate Std. Error t-value  p-value    
b:(Intercept) 1.15894518 0.08339821  13.896 0.005138 ** 
c:(Intercept) 0.24398009 0.01385086  17.615 0.003207 ** 
d:(Intercept) 1.81356338 0.05324814  34.059 0.000861 ***
e:(Intercept) 0.00261968 0.00020635  12.695 0.006148 ** 
---
Signif. codes:  
0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error:

 0.01758151 (2 degrees of freedom)

EDIT: Adding the log transformation leads to that the curve is not drawn anymore:

ggplot(test_data, aes(x = Conc, y = Response )) +
  geom_point() +
  stat_smooth(
              method = "drm",
              method.args = list(
                fct = LL.4()
              ),
              se = FALSE
  )+
  scale_x_continuous(trans="log10")

enter image description here

Upvotes: 1

Views: 991

Answers (2)

LBargie
LBargie

Reputation: 1

ggplot(test_data, aes(x = Conc, y = Response )) +
  geom_point() +
  geom_line(aes(y = predict(model_drm))) +
  scale_x_log10()

Upvotes: 0

G. Grothendieck
G. Grothendieck

Reputation: 270448

The X axis needs to be log10 scaled. Also we can reuse the calculation of the model already done.

ggplot(cbind(test_data, fit = fitted(model_drm)), aes(x = Conc, y = Response )) +
   geom_point() +
   geom_line(aes(y = fit)) +
   scale_x_continuous(trans = "log10")

screenshot

Upvotes: 1

Related Questions