bimodalTest {CHARGE} | R Documentation |
Performs a bimodal test and calcualtes Hartigan's statistic and p-value over a genomic region of interest using gene expression data set, the output from cvExpr.
bimodalTest(se, cvExpr, threshold = NULL)
se |
A SummarizedExperiment containing the normalised gene expression data. |
cvExpr |
The output from cvExpr. |
threshold |
Optional. The quantile threshold of genes to be used for clustering analaysis. Default is NULL. |
Performs a bimodal test and calculates Hartigan's dip test statistic for unimodality for a given gene expression data set. A bimodality coefficient value > 5/9 suggests bimodality and the closer the bimodality ratio is to 1, the more evenly distrubted the data set. The dip statistic and p-value can be used to determine if the region of interest is statistically significant.
The second part of the function returns the Z score means which can be used to visualise the denisty or distribution of the samples.
Returns a list containing the output from the bimodal test.
Benjamin Mayne
library(GenomicRanges) data(datExprs) chr21 <- GRanges("21:1-46709983") cvExpr.out <- cvExpr(se = datExprs, region = chr21) bimodalTest.out <- bimodalTest(se = datExprs, cvExpr = cvExpr.out, threshold = "25%")