FisherTest: Fisher's exact test for read counts on GRanges objects

In genomaths/MethylIT: Methylation Analysis Based on Signal Detection

Fisher's exact test for read counts on GRanges objects

Description

Given a GRanges object with the methylated and unmethylated read counts for control and treatment in its metacolumn, Fisher's exact test is performed for each cytosine site.

Usage

FisherTest(
  LR,
  count.col = c(1, 2),
  control.names = NULL,
  treatment.names = NULL,
  pooling.stat = "sum",
  tv.cut = NULL,
  hdiv.cut = NULL,
  hdiv.col = NULL,
  pAdjustMethod = "BH",
  pvalCutOff = 0.05,
  saveAll = FALSE,
  num.cores = 1L,
  tasks = 0L,
  verbose = FALSE,
  progressbar = TRUE,
  ...
)

Arguments

LR	A list of GRanges, a GRangesList, a CompressedGRangesList object, or an object from Methyl-IT downstream analyses: 'InfDiv' or 'pDMP' object. Each GRanges object from the list must have two columns: methylated (mC) and unmethylated (uC) counts. The name of each element from the list must coincide with a control or a treatment name.
count.col	2d-vector of integers with the indexes of the read count columns. If not given, then it is assumed that the methylated and unmethylated read counts are located in columns 1 and 2 of each GRanges metacolumns. If object LR is the output of Methyl-IT function estimateDivergence, then columns 1:4 are the read count columns: columns 1 and 2 are methylated and unmethylated read counts from the reference group, while columns 3 and 4 are methylated and unmethylated read counts from the treatment group, respectively. In this case, if the requested comparison is reference versus treatment, then no specification is needed for count.col. The comparison control versus treatment can be obtained by setting count.col = 3:4 and providing control.names and treatment.names.
control.names, treatment.names	Names/IDs of control and treatment samples, which must be included in the variable GR at the metacolumn. Default NULL. If provided the Fisher's exact test control versus treatment is performed. Default is NULL. If NULL, then it is assumed that each GRanges object in LR has four columns of counts. The first two columns correspond to the methylated and unmethylated counts from control/reference and the other two columns are the methylated and unmethylated counts from treatment, respectively.
pooling.stat	statistic used to estimate the methylation pool: row sum, row mean or row median of methylated and unmethylated read counts across individuals. If the number of control samples is greater than 2 and pooling.stat is not NULL, then they will pooled. The same for treatment. Otherwise, all the pairwise comparisons will be done.
tv.cut	A cutoff for the total variation distance (TVD; absolute value of methylation levels differences) estimated at each site/range as the difference of the group means of methylation levels. If tv.cut is provided, then sites/ranges k with |TV_k| < tv.cut are removed before performing the regression analysis. Its value must be NULL or a number 0 < tv.cut < 1.
hdiv.cut	An optional cutoff for the Hellinger divergence (hdiv). If the LR object derives from the previous application of function estimateDivergence, then a column with the hdiv values is provided. If combined with tv.cut, this permits a more effective filtering of the signal from the noise. Default is NULL.
hdiv.col	Optional. Columns where hdiv values are located in each GRanges object from LR. It must be provided if together with hdiv.cut. Default is NULL.
pAdjustMethod	method used to adjust the results; default: BH
pvalCutOff	cutoff used then a p-value adjustment is performed
saveAll	if TRUE all the temporal results are returned
num.cores	The number of cores to use, i.e. at most how many child processes will be run simultaneously (see bpapply function from BiocParallel).
tasks	integer(1). The number of tasks per job. value must be a scalar integer >= 0L. In this documentation a job is defined as a single call to a function, such as bplapply, bpmapply etc. A task is the division of the X argument into chunks. When tasks == 0 (default), X is divided as evenly as possible over the number of workers (see MulticoreParam from BiocParallel package).
verbose	if TRUE, prints the function log to stdout
progressbar	logical(1). Enable progress bar
...	Additional parameters for function uniqueGRanges.

Details

Samples from each group are pooled according to the statistic selected (see parameter pooling.stat) and a unique GRanges object is created with the methylated and unmethylated read counts for each group (control and treatment) in the metacolumn. So, a contingency table can be built for range from GRanges object.

Value

The input GRanges object with the columns of Fisher's exact test p-value, total variation (difference of methylation levels), and p-value adjustment.

See Also

rmstGR

Examples

## Get a dataset of Hellinger divergency of methylation levels
## from the package
data(HD)

### --- To get the read counts
hd <- lapply(HD, function(hd) {
hd = hd[1:10,3:4]
colnames(mcols(hd)) <- c('mC', 'uC')
return(hd)
})

FisherTest(LR = hd, pooling.stat = NULL, control.names = 'C1',
treatment.names = 'T1', pAdjustMethod='BH', pvalCutOff = 0.05,
num.cores = 1L, verbose=FALSE)

genomaths/MethylIT documentation built on Feb. 3, 2024, 1:24 a.m.