PredictABEL (version 1.2-4)

ORmultivariate: Function to obtain multivariate odds ratios from a logistic regression model.

Description

The function estimates multivariate (adjusted) odds ratios (ORs) with 95% confidence intervals (CIs) for all the genetic and non-genetic variables in the risk model.

Usage

ORmultivariate(riskModel, filename)

Arguments

riskModel

Name of logistic regression model that can be fitted using the function fitLogRegModel.

filename

Name of the output file in which the multivariate ORs will be saved. If no directory is specified, the file is saved in the working directory as a txt file. When filename is not specified, the output is not saved.

Value

The function returns:

Predictors Summary

OR with 95% CI and corresponding p-values for each predictor in the model

Brier Score

Brier score

Nagelkerke Index

Nagelkerke's \(R^2\) value

Details

The function requires that first a logistic regression model is fitted either by using GLM function or the function fitLogRegModel. In addition to the multivariate ORs, the function returns summary statistics of model performance, namely the Brier score and the Nagelkerke's \(R^2\) value. The Brier score quantifies the accuracy of risk predictions by comparing predicted risks with observed outcomes at individual level (where outcome values are either 0 or 1). The Nagelkerke's \(R^2\) value indicates the percentage of variation of the outcome explained by the predictors in the model.

References

Brier GW. Verification of forecasts expressed in terms of probability. Monthly weather review 1950;78:1-3.

Nagelkerke NJ. A note on a general definition of the coefficient of determination. Biometrika 1991;78:691-692.

See Also

fitLogRegModel

Examples

# NOT RUN {
# specify dataset with outcome and predictor variables
data(ExampleData)
# specify column number of outcome variable
cOutcome <- 2
# specify column numbers of non-genetic predictors
cNonGenPred <- c(3:10)
# specify column numbers of non-genetic predictors that are categorical
cNonGenPredCat <- c(6:8)
# specify column numbers of genetic predictors
cGenPred <- c(11,13:16)
# specify column numbers of genetic predictors that are categorical
cGenPredCat <- c(0)

# fit logistic regression model
riskmodel <- fitLogRegModel(data=ExampleData, cOutcome=cOutcome,
cNonGenPreds=cNonGenPred, cNonGenPredsCat=cNonGenPredCat,
cGenPreds=cGenPred, cGenPredsCat=cGenPredCat)

# obtain multivariate OR(95% CI) for all predictors of the fitted model
ORmultivariate(riskModel=riskmodel)
# }