getRegionScores.RdReturns p-values and Z-scores for CpGs in a set of regions, compared to a local null background distribution.
getRegionScores(ranks, regions, chromDictObj)getPCRanks output data.frame, only necessary if chromDictObj=NULL
A three-column dataframe containing a set of regions to test. Columns = chrom, start, end.
chromDict() output object, recommended input instead of ranks.
Returns a data.frame with significance scores for all input regions.
ranks <- getPCRanks(eigen, IDs = c("trt", "ctl"), PC = 1)
DMLs <- addRanks(ranks)
# data.frame of regions to test:
regions <- data.frame(chr=c("chr3", "chr3", "chr1"),
s=c(4920450, 3961576, 300000),
e=c(4923267, 3963805, 302900),
ID=c("Hypo-DMR", "partial Hyper-DMR", "random"))
getRegionScores(DMLs, regions)
#> Creating chromDict Object. If you plan to run the getRegionScores() function multiple times, it is strongly recommended to generate a chromDict Object with the chromDict() function, and specify it with getRegionScores(..., chromDictObj=OBJECT). This is the most computationally intensive part of most PCBS functions, and only needs to be done once per dataset.
#> Getting Background Sites...
#> Checking significance at 3 sites...
#> chrom start end feature meanPC nCpG Z
#> 1 chr3 4920450 4923267 Hypo-DMR -0.020963435 24 -4.5222576
#> 2 chr3 3961576 3963805 partial Hyper-DMR 0.008092709 8 1.8522800
#> 3 chr1 300000 302900 random -0.003612135 8 -0.9741492
#> p
#> 1 9.146579e-08
#> 2 1.373174e-06
#> 3 9.043174e-04