TestAssociations: Test associations between phenotype and RNA editing levels.

In rnaEditr: Statistical analysis of RNA editing sites and hyper-editing regions

Description

A wrapper function to test associations between phenotype and RNA editing levels in single-site analysis or summarized RNA editing levels in region-based analysis.

Usage

TestAssociations(
  rnaEdit_df,
  pheno_df,
  responses_char,
  covariates_char = NULL,
  respType = c("binary", "continuous", "survival"),
  progressBar = "time",
  orderByPval = TRUE
)

Arguments

rnaEdit_df	A data frame with class rnaEdit_df, which is a output from function CreateEditingTable() or function SummarizeAllRegions(). This data frame should include RNA editing level values, with row names as site IDs or region IDs, and column names as sample IDs.
pheno_df	A data frame with phenotype and covariates, which should include all the samples in rnaEdit_df. Please make sure the input pheno_df has the variable named "sample" to indicate sample IDs.
responses_char	A character vector of names of response variables in pheno_df. When respType is set as "survival", responses_char should have length 2. The first element must be the name of the variable with following up time, and the second element must be status indicator. Status indicator should be coded as 0/1(1=death), TRUE/FALSE(TRUE=death), or 1/2(death). Please make sure variable names are in this order. We have not tested this code on interval-censored data; use at your own risk. See Surv for more details.
covariates_char	A character vector of names of covariate variables in pheno_df.
respType	Type of outcome. Defaults to "binary".
progressBar	Name of the progress bar to use. There are currently five types of progress bars: "time", "none", "text", "tk", and "win". Defaults to "time". See create_progress_bar for more details.
orderByPval	Sort co-edited regions by model p-value or not? Defaults to TRUE.

Value

A data frame with locations of the genomic regions or sites (seqnames, start, end, width), test statistics (estimate, stdErr or coef, exp_coef, se_coef), pValue and false discovery rate (fdr).

See Also

TransformToGR, AllCloseByRegions, AllCoeditedRegions, CreateEditingTable, SummarizeAllRegions, AnnotateResults

Examples

  data(rnaedit_df)
  
  genes_gr <- TransformToGR(
    genes_char = c("PHACTR4", "CCR5", "METTL7A"),
    type = "symbol",
    genome = "hg19"
  )
  
  exm_regions <- AllCoeditedRegions(
    regions_gr = genes_gr,
    rnaEditMatrix = rnaedit_df,
    output = "GRanges",
    method = "spearman"
  )
  
  sum_regions <- SummarizeAllRegions(
    regions_gr = exm_regions,
    rnaEditMatrix = rnaedit_df,
    selectMethod = MaxSites
  )
  
  exm_pheno <- readRDS(
    system.file(
    "extdata",
    "pheno_df.RDS",
    package = 'rnaEditr',
    mustWork = TRUE
    )
  )
  
  TestAssociations(
    rnaEdit_df = sum_regions,
    pheno_df = exm_pheno,
    responses_char = "sample_type",
    covariates_char = NULL,
    respType = "binary"
  )

rnaEditr documentation built on Nov. 8, 2020, 8:26 p.m.