ratePvals {INSPEcT} | R Documentation |
Retrieve a single p-value for each rate
This method is used to retrieve all the p-values combined with Brown's method that combines the results of the
log likelihood ratio test results for all pairs tested for each rate and all genes. P-values will change
according to the threshold set for the chi-squared test because it influences the model that will be taken into
consideration to perform log likelihood ratio tests. To have a sense of the best parameter to choose, a sythetic
data-set can be built and tested (makeSimModel
, makeSimDataset
)
In case 'aic' has been selected via modelSelection
method,
this method assigns the chi-squared test result of the model selected by AIC
to the respective variable rates
ratePvals(object, cTsh = NULL) ## S4 method for signature 'INSPEcT_model' ratePvals(object, cTsh = NULL) ## S4 method for signature 'INSPEcT' ratePvals(object, cTsh = NULL)
object |
An object of class INSPEcT or INSPEcT_model |
cTsh |
A numeric representing the threshold for the chi-squared test to consider a model as valid |
ratePvlas retrieve a single p-value for each rate thanks to multiple log likelihood tests performed on nested models that has a chi-squared test below the selected threshold. Among the many p-values that log likelihood ratio test calculate, a single p-value is obtaied applying Brown's method for combining dependent p-values.
A matrix containing p-values calculated for each rate
data('mycerIds10', package='INSPEcT') ratePvals(mycerIds10) # calculate agin the p-values with Brown with a different threshold # for considering a model valid for the log likelihood ratio test ratePvals(mycerIds10, cTsh=.2) # Set permaenently the chi-squared threshold at .2 for mycerIds10 object thresholds(mycerIds10)$chisquare <- .2