maxR {BIGL}  R Documentation 
maxR
computes maxR statistics for each offaxis dose
combination given the data provided. It provides a summary with results
indicating whether a given point is estimated to be synergetic or
antagonistic. These can be based either on normal approximation or a
fully bootstrapped distribution of the statistics.
maxR(
data_off,
fitResult,
transforms = fitResult$transforms,
null_model = c("loewe", "hsa", "bliss", "loewe2"),
R,
CP,
reps,
nested_bootstrap = FALSE,
B.B = NULL,
cutoff = 0.95,
cl = NULL,
B.CP = NULL,
method = c("equal", "model", "unequal"),
bootStraps,
idUnique,
n1,
doseGridOff,
transFun,
invTransFun,
...
)
data_off 
data frame with off axis information 
fitResult 
Monotherapy (onaxis) model fit, e.g. produced by

transforms 
Transformation functions. If nonnull, 
null_model 
Specified null model for the expected response surface.
Currently, allowed options are 
R 
Numeric vector containing mean deviation of predicted response
surface from the observed one at each of the offaxis points. If missing,
it will be calculated automatically from output of

CP 
Prediction covariance matrix. If not specified, it will be estimated
by bootstrap using 
reps 
Numeric vector containing number of replicates for each offaxis
dose combination. If missing, it will be calculated automatically from output
of 
nested_bootstrap 
When statistics are calculated, if

B.B 
Number of iterations to use in bootstrapping null distribution for either meanR or maxR statistics. 
cutoff 
Cutoff to use in maxR procedure for declaring nonadditivity (default is 0.95). 
cl 
If parallel computations are desired, 
B.CP 
Number of bootstrap iterations to use for CP matrix estimation 
method 
What assumption should be used for the variance of on and
offaxis points. This argument can take one of the values from

bootStraps 
precomputed bootstrap objects 
idUnique 
unique combinations of onaxis points, a character vector 
n1 
the number of offaxis points 
doseGridOff 
dose grid for offaxis points 
transFun 
the transformation and inverse transformation functions for the variance 
invTransFun 
the transformation and inverse transformation functions for the variance 
... 
Further arguments that will be later passed to

This function returns a maxR
object with estimates for the
maxR statistical test. maxR
object is essentially a list with
appropriately named elements.
In particular, maxR
object contains "Ymean"
element which is
a summary table of maxR test results for each dose combination. This table
contains mean deviation from the predicted surface, normalized deviation
("absR"
) as well as a statistical call whether this deviation is
significant. Distributional information on which these calls are made can
be retrieved from the attributes of the "Ymean"
dataframe.
Also, maxR
object contains "Call"
element which indicates the
general direction of the deviation of the observed surface from the null.
This call is based on the strongest local deviation in the "Ymean"
table. 4 values are available here: "Syn"
, "Ant"
,
"None"
, "Undefined"
. If one compound acts as an agonist while
another one is an antagonist, then a deviation from the null is classified
as "Undefined"
. If both compounds act in the same direction, then a
stronger than individual effect is classified as synergy while a weaker
effect would be classified as antagonism.