GWAS object — GWAS • genepi.utils

A GWAS object is a container for vectors of GWAS data, a correlation matrix, and meta-data regarding quality control procedures applied at the point of object creation / data import.

Usage

GWAS(
  dat,
  map = "default",
  drop = FALSE,
  fill = FALSE,
  fill_rsid = FALSE,
  missing_rsid = "fill_CHR:BP",
  parallel_cores = parallel::detectCores(),
  dbsnp_dir = genepi.utils::which_dbsnp_directory(),
  filters = list(beta_invalid = "!is.infinite(beta) & abs(beta) < 20", eaf_invalid =
    "eaf > 0 & eaf < 1", p_invalid = "!is.infinite(p)", se_invalid = "!is.infinite(se)",
    alleles_invalid = "!is.na(ea) & !is.na(oa)", chr_missing = "!is.na(chr)", bp_missing
    = "!is.na(bp)", beta_missing = "!is.na(beta)", se_missing = "!is.na(se)", p_missing =
    "!is.na(p)", eaf_missing = "!is.na(eaf)"),
  reference = NULL,
  ref_map = NULL,
  verbose = TRUE,
  ...
)

Arguments

dat: a valid string file path to be read by data.table::fread or a data.table::data.table object; the GWAS data source
map: a valid input to the ColumnMap class constructor (a predefined map string id, a named list or character vector, or a ColumMap object)
drop: a logical, whether to drop data source columns not in the column map
fill: a logical, whether to add (NAs) missing columns present in the column map but not present in the data source
fill_rsid: either FALSE or a valid argument for the chrpos_to_rsid build argument, e.g. "b37_dbsnp156"
missing_rsid: a string, how to handle missing rsids: one of "fill_CHR:BP", "fill_CHR:BP_OA_EA", "overwrite_CHR:BP", "overwrite_CHR:BP:OA:EA", "none", or "leave"
parallel_cores: an integer, number of cores to used for RSID mapping, default is maximum machine cores
dbsnp_dir: path to the dbsnp directory of fst files see chrpos_to_rsid dbsnp_dir argument
filters: a list of named strings, each to be evaluated as an expression to filter the data during the quality control steps (above)
reference: a valid string file path to be read by data.table::fread or a data.table::data.table object; the reference data
ref_map: a valid input to the ColumnMap class constructor (a predefined map id (a string), a named list or character vector, or a ColumMap object) defining at least columns rsid (or chr, bp), ea, oa and eaf.
verbose: a logical, whether to print details
...: variable capture to be passed to the constructor, e.g. individual vectors for the slots, rather that dat

Value

an S7 class genepi.utils::GWAS object

Slots

rsid: character, variant ID - usually in rs12345 format, however this can be changed with the missing_rsid argument
chr: character, chromosome identifier
bp: integer, base position
ea: character, effect allele
oa: character, other allele
eaf: numeric, effect allele frequency
beta: numeric, effect size
se: numeric, effect size standard error
p: numeric, p-value
n: integer, total number of samples
ncase: integer, number of cases
strand: character, the strand + or -
imputed: logical, whether imputed
info: numeric, the info score
q: numeric, the Q statistic for meta analysis results
q_p: numeric, the Q statistic P-value
i2: numeric, the I2 statistic
proxy_rsid: character, proxy variant ID
proxy_chr: character, proxy chromosome identifier
proxy_bp: integer, proxy base position
proxy_ea: character, proxy effect allele
proxy_oa: character, proxy other allele
proxy_eaf: numeric, proxy effect allele frequency
proxy_r2: numeric, proxy r2 with rsid
trait: character, the GWAS trait
id: character, the GWAS identifier
source: character, data source; either the file path, or "data.table" if loaded directly
correlation: matrix, a correlation matrix of signed R values between variants
map: ColumnMap, a mapping of class ColumnMap
qc: list, a named list of filters; name is the filter expression and value is an integer vector of rows that fail the filter