Build a column rename map from per-cohort manifest values

Constructs a named character vector mapping non-standard column names (as they appear in the raw summary statistics file) to harmonised names. Each col_* argument corresponds to a single harmonised column. Pass the cohort-specific raw column name (a string) to override the default mapping, or leave as NA to skip.

The returned vector is suitable for the column_map argument of harmonize_sumstats_headers().

Usage

build_column_map(
  col_chr = NA_character_,
  col_pos = NA_character_,
  col_rsid = NA_character_,
  col_effect_allele = NA_character_,
  col_other_allele = NA_character_,
  col_beta = NA_character_,
  col_se = NA_character_,
  col_p = NA_character_,
  col_eaf = NA_character_,
  col_n = NA_character_,
  col_n_cases = NA_character_,
  col_n_controls = NA_character_
)

Arguments

col_chr

Column name mapping to CHR.

col_pos

Column name mapping to POS.

col_rsid

Column name mapping to RSID.

col_effect_allele

Column name mapping to EffectAllele.

col_other_allele

Column name mapping to OtherAllele.

col_beta

Column name mapping to B.

col_se

Column name mapping to SE.

col_p

Column name mapping to P.

col_eaf

Column name mapping to EFF_ALL_FREQ.

col_n

Column name mapping to N.

col_n_cases

Column name mapping to N_CASES.

col_n_controls

Column name mapping to N_CONTROLS.

Value

A named character vector where names are source column names and values are target harmonised names, or NULL if all arguments are NA.

Examples

build_column_map(col_eaf = "MY_FREQ", col_beta = "BETA_VAL")
#>       BETA_VAL        MY_FREQ 
#>            "B" "EFF_ALL_FREQ" 
# c("MY_FREQ" = "EFF_ALL_FREQ", "BETA_VAL" = "B")

build_column_map()
#> NULL
# NULL