computeRSquared: Compute R and R^2 between rows of two SummarizedExperiment...
In AffiXcan: A Functional Approach To Impute Genetically Regulated Expression

Description Usage Arguments Value Examples

Compute R and R^2 between rows of two SummarizedExperiment assays

computeRSquared(
  realExpr,
  imputedExpr,
  assay,
  testingSamples = NULL,
  BPPARAM = bpparam()
)

`realExpr`	A SummarizedExperiment object containing expression data
`imputedExpr`	The returning object of affiXcanImpute()
`assay`	A string with the name of the object in SummarizedExperiment::assays(realExpr) that contains expression values
`testingSamples`	A vector of strings. The identifiers of the samples that have to be considered by the function; default is NULL; if is.null(testingSamples)==TRUE then no filtering is performed
`BPPARAM`	A BiocParallelParam object. Default is bpparam(). For details on BiocParallelParam virtual base class see browseVignettes("BiocParallel")

A list of lists; inner lists are named after the rows for which the correlation between realExpr and imputedExpr have been computed; inner lists contain two objects:

rho: the pearson's correlation coefficient (R) between the real expression values and the imputed GReX for the cross-validation i on the expressed gene y, computed with cor()
rho.sq: the coefficient of determination (R^2) between the real expression values and the imputed GReX for the cross-validation i on the expressed gene y, computed as pearson^2
cor.test.p.val: the p-value of the cor.test() between the real expression values and the imputed GReX for the cross-validation i on the expressed gene y

if (interactive()) {
trainingTbaPaths <- system.file("extdata","training.tba.toydata.rds",
package="AffiXcan")

data(exprMatrix)
data(regionAssoc)
data(trainingCovariates)

assay <- "values"

training <- affiXcanTrain(exprMatrix=exprMatrix, assay=assay,
tbaPaths=trainingTbaPaths, regionAssoc=regionAssoc, cov=trainingCovariates,
varExplained=80, scale=TRUE)

imputedExpr <- affiXcanImpute(tbaPaths=trainingTbaPaths,
affiXcanTraining=training, scale=TRUE)
realExpr <- exprMatrix

correlation <- computeRSquared(realExpr, imputedExpr, assay)
}