evaRIVER: Evaluation of RIVER
In RIVER: R package for RIVER (RNA-Informed Variant Effect on Regulation)

Description Usage Arguments Value Warning Author(s) See Also Examples

evaRIVER trains RIVER by holding out a list of individual and gene pairs having same rare variants for evaluation, computes test posterior probabilities of FR for 1st individual, and compares them with outlier status of 2nd individual from the list.

1
2
3

evaRIVER(dataInput, pseudoc = 50, theta_init = matrix(c(0.99, 0.01, 0.3,
  0.7), nrow = 2), costs = c(100, 10, 1, 0.1, 0.01, 0.001, 1e-04),
  verbose = FALSE)

`dataInput`	An object of ExpressionSet class which contains input data required for all functions in RIVER including genomic features, outlier status, and N2 pairs.
`pseudoc`	Pseudo count.
`theta_init`	Initial values of theta.
`costs`	Candidate penalty parameter values for L2-regularized logistic regression.
`verbose`	Logical option for showing extra information on progress.

A list which contains two AUC values from RIVER and GAM, computed specificities and sensitivities from two models, and P-value of comparing the two AUC values.

A vector of candidate penalty values make glmnet faster than to input a single penalty value

Yungil Kim, ipw012@gmail.com

cv.glmnet, predict, integratedEM, testPosteriors, getData, exprs

1
2
3

dataInput <- getData(filename=system.file("extdata", "simulation_RIVER.gz",
        package = "RIVER"), ZscoreThrd=1.5)
evaROC <- evaRIVER(dataInput, verbose=TRUE)