summaryROCPlot: Generate a plot with a summary ROC curve

In MetaIntegrator: Meta-Analysis of Gene Expression Data

Description

Generate a plot with a summary ROC curve

Usage

summaryROCPlot(metaObject, filterObject, bootstrapReps = 500,
  orderByAUC = TRUE, alphaBetaPlots = TRUE)

Arguments

metaObject	a Meta object which must have the $originalData populated
filterObject	a MetaFilter object containing the signature genes that will be used for calculating the score
bootstrapReps	number of bootstrap simulations to run for confidence interval on summary ROC
orderByAUC	if TRUE, then order legend by summary AUC. Otherwise, use default ordering.
alphaBetaPlots	if TRUE, then draw forest plots of alpha and beta. If false, suppress plotting.

Value

Generates a ROC plot for all datasets

Author(s)

Timothy E. Sweeney

Examples

## Not run: 
summaryROCPlot(tinyMetaObject,filterObject = 
   tinyMetaObject$filterResults$pValueFDR0.05_es0_nStudies1_looaTRUE_hetero0)

## End(Not run)

Example output

 Whole Blood Study 1 		Bootstrapping...	
 Whole Blood Study 2 		Bootstrapping...	
 PBMC Study 1 		Bootstrapping...	
                      N tstar_alpha   SE_alpha tstar_beta    SE_beta
Whole Blood Study 1  49    2.290973 0.07439752 -0.2658663 0.05670437
Whole Blood Study 2 142    2.755875 0.02174051 -0.2138215 0.01161228
PBMC Study 1        115    4.403337 0.03749007 -0.3617431 0.01628389

 Whole Blood Study 1 		Bootstrapping...	
 Whole Blood Study 2 		Bootstrapping...	
 PBMC Study 1 		Bootstrapping...	
                      N tstar_alpha   SE_alpha tstar_beta    SE_beta
Whole Blood Study 1  49    2.286690 0.07326695 -0.2633483 0.05353488
Whole Blood Study 2 142    2.755560 0.02103828 -0.2136221 0.01170006
PBMC Study 1        115    4.402778 0.03950574 -0.3614891 0.01730615
Random-effects meta-analysis
Call: rmeta::meta.summaries(d = tstar_alpha, se = SE_alpha * sqrt(N), 
    method = method)
Summary effect=3.15   95% CI (2, 4.31)
Estimated heterogeneity variance: 0.88  p= 0.001 
Random-effects meta-analysis
Call: rmeta::meta.summaries(d = tstar_beta, se = SE_beta * sqrt(N), 
    method = method)
Summary effect=-0.267   95% CI (-0.476, -0.0573)
Estimated heterogeneity variance: 0  p= 0.816 
Summary ROC 
At optimal cutoff, summary sensitivity = 78.33%, summary specificity = 88.9%
At summary sensitivity = 95%, summary specificity = 0.313%                     sens  spec   PPV   NPV   acc
Whole Blood Study 1 0.758 0.875 0.926 0.636 0.796
Whole Blood Study 2 0.769 0.844 0.857 0.750 0.803
PBMC Study 1        0.903 0.833 0.979 0.500 0.896

MetaIntegrator documentation built on March 26, 2020, 6:29 p.m.