braidrm

The goal of braidrm is to to make the best combination analysis available more robust, more accessible, and easier to use, so that drug combinations can be understood more completely and new therapies can be discovered more quickly.

Example

This example shows how to fit a BRAID response surface to data, and print a summary of the resulting fit.

library(braidrm)

# Fit a basic braid surface
braidFit <- braidrm(measure ~ concA + concB, synergisticExample,
                    model = "kappa2", getCIs=TRUE)

summary(braidFit)
#> Call:
#> braidrm.formula(formula = measure ~ concA + concB, data = synergisticExample, 
#>     model = "kappa2", getCIs = TRUE)
#> 
#>            Lo     Est     Hi
#> IDMA   0.8927  1.0398 1.1770
#> IDMB   0.8887  1.0259 1.1859
#> na     2.3640  2.9116 3.5669
#> nb     2.0537  2.4990 3.3016
#> kappa  1.7074  2.1258 2.5839
#> E0    -0.0766 -0.0300 0.0227
#> EfA    0.9281  1.0080 1.0245
#> EfB    0.9169  0.9848 1.0240
#> Ef         NA  1.0080     NA