GRAB (version 0.2.3)

GRAB.WtCoxG: WtCoxG method in GRAB package

Description

WtCoxG is an accurate, powerful, and computationally efficient Cox-based approach to perform genome-wide time-to-event data analyses in study cohorts with case ascertainment.

Usage

GRAB.WtCoxG()

Value

No return value, called for side effects (prints information about the WtCoxG method to the console).

Details

Additional arguments in GRAB.NullModel():

• RefAfFile: A character string specifying a reference allele frequency file, which is a csv file (with a header) and includes columns of CHROM, POS, ID, REF, ALT, AF_ref, and AN_ref.

• OutputFile: A character string specifying the output file name.

• SampleIDColumn: A character string specifying the column name in the input data that contains sample IDs.

• SurvTimeColumn: A character string specifying the column name in the input data that contains survival time information.

• IndicatorColumn: A character string specifying the column name in the input data that indicates case-control status (should be 0 for controls and 1 for cases).

Additional arguments in list control in GRAB.NullModel():

• RefPrevalence: A numeric value specifying the population-level disease prevalence used for weighting in the analysis.

• SNPnum: Minimum number of SNPs. Default is 1e4.

Additional arguments in list control in GRAB.Marker():

• cutoff: A numeric value specifying the batch effect p-value cutoff for method selection of an association test. Default is 0.1.

Examples

# Step0&1: fit a null model and estimate parameters according to batch effect p-values
PhenoFile <- system.file("extdata", "simuPHENO.txt", package = "GRAB")
PhenoData <- data.table::fread(PhenoFile, header = TRUE)
SparseGRMFile <- system.file("extdata", "SparseGRM.txt", package = "GRAB")

GenoFile <- system.file("extdata", "simuPLINK.bed", package = "GRAB")
RefAfFile <- system.file("extdata", "simuRefAf.txt", package = "GRAB")
RefPrevalence <- 0.1 # population-level disease prevalence

OutputDir <- tempdir()
OutputStep1 <- file.path(OutputDir, "WtCoxG_step1_out.txt")
OutputStep2 <- file.path(OutputDir, "WtCoxG_step2_out.txt")

obj.WtCoxG <- GRAB.NullModel(
  formula = survival::Surv(SurvTime, SurvEvent) ~ AGE + GENDER,
  data = PhenoData,
  subjData = PhenoData$IID,
  method = "WtCoxG",
  traitType = "time-to-event",
  GenoFile = GenoFile,
  SparseGRMFile = SparseGRMFile,
  control = list(
    AlleleOrder = "ref-first", 
    AllMarkers = TRUE, 
    RefPrevalence = RefPrevalence,
    SNPnum = 1000  # the minimum number of SNPs that TestforBatchEffect() needs
  ), 
  RefAfFile = RefAfFile,
  OutputFile = OutputStep1,
  SampleIDColumn = "IID",
  SurvTimeColumn = "SurvTime",
  IndicatorColumn = "SurvEvent"
)

resultStep1 <- data.table::fread(OutputStep1)
resultStep1[, c("CHROM", "POS", "pvalue_bat")]

# Step2: conduct association testing
GRAB.Marker(
  objNull = obj.WtCoxG,
  GenoFile = GenoFile,
  OutputFile = OutputStep2,
  control = list(
    AlleleOrder = "ref-first", 
    AllMarkers = TRUE,
    cutoff = 0.1, 
    nMarkersEachChunk = 5000
  )
)

resultStep2 <- data.table::fread(OutputStep2)
resultStep2[, c("CHROM", "POS", "WtCoxG.noext", "WtCoxG.ext")]