runPermutation: Run all permutations on the specified multi-generational...

In adeschen/methylInheritance: Permutation-Based Analysis associating Conserved Differentially Methylated Elements Across Multiple Generations to a Treatment Effect

Description

Run a permutation analysis, based on Monte Carlo sampling, for testing the hypothesis that the number of conserved differentially methylated elements (sites, tiles or both), between several generations, is associated to an effect inherited from a treatment and that stochastic effect can be dismissed.

The multi-generational dataset or the name of the RDS file that contains the dataset can be used as input.

The observation analysis can also be run (optional). All permutation results are saved in RDS files.

Usage

runPermutation(
  methylKitData,
  type = c("both", "sites", "tiles"),
  outputDir = "output",
  runObservationAnalysis = TRUE,
  nbrPermutations = 1000,
  nbrCores = 1,
  nbrCoresDiffMeth = 1,
  minReads = 10,
  minMethDiff = 10,
  qvalue = 0.01,
  maxPercReads = 99.9,
  destrand = FALSE,
  minCovBasesForTiles = 0,
  tileSize = 1000,
  stepSize = 1000,
  vSeed = -1,
  restartCalculation = FALSE,
  saveInfoByGeneration = FALSE
)

Arguments

methylKitData	a list of methylRawList entries or the name of the RDS file containing the list. Each methylRawList entry must contain all the methylRaw entries related to one generation (first entry = first generation, second entry = second generation, etc..). The number of generations must correspond to the number of entries in the methylKitData. At least 2 generations must be present to make a permutation analysis. More information can be found in the methylKit package.
type	One of the "sites","tiles" or "both" strings. Specifies the type of differentially methylated elements should be returned. For retrieving differentially methylated bases type="sites"; for differentially methylated regions type="tiles". Default: "both".
outputDir	a string, the name of the directory that will contain the results of the permutation. If the directory does not exist, it will be created. Default: "output".
runObservationAnalysis	a logical, when runObservationAnalysis = TRUE, a CpG analysis on the observed dataset is done. Default: TRUE.
nbrPermutations,	a positive integer, the total number of permutations that is going to be done. Default: 1000.
nbrCores	a positive integer, the number of cores to use when processing the analysis. Default: 1 and always 1 for Windows.
nbrCoresDiffMeth	a positive integer, the number of cores to use for parallel differential methylation calculations. The parameter is used for both sites and tiles analysis. The parameter corresponds to the num.cores parameter in the package methylKit. Default: 1 and always 1 for Windows.
minReads	a positive integer Bases and regions having lower coverage than this count are discarded. The parameter corresponds to the lo.count parameter in the package methylKit.
minMethDiff	a positive double between [0,100], the absolute value of methylation percentage change between cases and controls. The parameter corresponds to the difference parameter in the methylKit package. Default: 10.
qvalue	a positive double between [0,1], the cutoff for qvalue of differential methylation statistics. Default: 0.01.
maxPercReads	a double between [0,100], the percentile of read counts that is going to be used as an upper cutoff. Bases or regions having higher coverage than this percentile are discarded. The parameter is used for both CpG sites and tiles analysis. The parameter corresponds to the hi.perc parameter in the package methylKit. Default: 99.9.
destrand	a logical, when TRUE will merge reads on both strands of a CpG dinucleotide to provide better coverage. Only advised when looking at CpG methylation. The parameter is used for both CpG sites and tiles analysis. Default: FALSE.
minCovBasesForTiles	a non-negative integer, the minimum number of bases to be covered in a given tiling window. The parameter corresponds to the cov.bases parameter in the package methylKit. Only used when doingTiles = TRUE. Default: 0.
tileSize	a positive integer, the size of the tiling window. The parameter corresponds to the win.size parameter in the package methylKit. Only used when doingTiles = TRUE. Default: 1000.
stepSize	a positive integer, the step size of tiling windows. The parameter corresponds to the stepSize parameter in the package methylKit. Only used when doingTiles = TRUE. Default: 1000.
vSeed	a integer, a seed used when reproducible results are needed. When a value inferior or equal to zero is given, a random integer is used. Default: -1.
restartCalculation	a logical, when TRUE, only permutations that don't have an associated RDS result file are run. Useful to restart a permutation analysis that has been interrupted. Beware that the parameters have to be identical except for this one.
saveInfoByGeneration	a logical, when TRUE, the information about differentially methylated sites and tiles for each generation is saved in a RDS file. The information is saved in a different file for each permutation. The files are saved in the directory specified by the outputDir parameter.

Value

0.

Author(s)

Astrid Deschenes, Pascal Belleau

See Also

mergePermutationAndObservation for detail description, in the Value section, of the methylInheritanceAllResults object as well as its PERMUTATION section.

Examples

## Load methylKit information
data(samplesForTransgenerationalAnalysis)

## Run a permutation analysis using the methylKit dataset
## A real analysis would require a much higher number of permutations
runPermutation(methylKitData = samplesForTransgenerationalAnalysis,
    outputDir = "test_01", runObservationAnalysis = FALSE, type = "sites",
    nbrPermutations = 2, vSeed = 221)

## Get results
results_01 <- loadAllRDSResults(analysisResultsDir = NULL,
    permutationResultsDir = "test_01", doingSites = TRUE,
    doingTiles = FALSE)

## Remove results directory
if (dir.exists("test_01")) {
    unlink("test_01", recursive = TRUE, force = TRUE)
}

## Path to a methylKit RDS file
methylFile <- system.file("extdata", "methylObj_001.RDS",
    package = "methylInheritance")

## Run a permutation analysis using RDS file name
## A real analysis would require a much higher number of permutations
runPermutation(methylKitData = methylFile, type = "tiles",
    outputDir = "test_02", nbrPermutations = 2, minCovBasesForTiles = 10,
    vSeed = 2001)

## Get results
results_02 <- loadAllRDSResults(analysisResultsDir = NULL,
    permutationResultsDir = "test_02", doingSites = FALSE,
    doingTiles = TRUE)

## Remove results directory
if (dir.exists("test_02")) {
    unlink("test_02", recursive = TRUE, force = TRUE)
}

adeschen/methylInheritance documentation built on April 21, 2021, 9:45 a.m.