data-raw/snp_annotations.R

In tbilab/meToolkit: Build and Customize Phewas Multimorbidity Explorer Apps.

## code to prepare `snp_annotations` dataset goes here
library(tidyverse)

clean_text_field <- . %>%
  str_replace_all('&', ', ') %>%
  str_replace_all('_', ' ') %>%
  tolower()

snp_annotations <- read_csv(
  here::here('data/full_exome_annotations.csv'),
  col_names = TRUE,
  cols(
    snp = col_character(),
    chr = col_character(),
    gene = col_character(),
    rsid = col_character(),
    Consequence = col_character(),
    IMPACT = col_character(),
    SYMBOL = col_character(),
    Gene = col_character(),
    HGVSp = col_character(),
    Existing_variation = col_character(),
    ExAC_MAF = col_character(),
    CLIN_SIG = col_character()
  )) %>%
  rename(
    name = snp,
    rsid = rsid,
    consequence = Consequence,
    impact = IMPACT,
    symbol = SYMBOL,
    gene_id = Gene,
    existing_variant = Existing_variation,
    clinical_sig = CLIN_SIG
  ) %>%
  separate(
    ExAC_MAF,
    into = c('minor_allele', 'ExAC_maf'),
    sep = ":"
  ) %>%
  mutate(
    ExAC_maf = str_remove_all(ExAC_maf, '&[A,T,C,G,U]+'),
    clinical_sig = clean_text_field(clinical_sig),
    consequence = clean_text_field(consequence),
    impact = clean_text_field(impact)
  ) %>%
  select(
    rsid,
    name,
    chr,
    gene,
    minor_allele,
    ExAC_maf,
    clinical_sig,
    consequence,
    impact,
    symbol,
    gene_id,
    HGVSp
  ) %>%
  bind_rows( # Add a fake snp for use for simulated data
    tibble(
      rsid = 'rs123456',
      name = 'fake snp',
      chr = '1',
      gene = 'sample gene',
      minor_allele = 'U',
      ExAC_maf = '0.123456',
      clinical_sig = 'benign',
      consequence = 'nothing',
      impact = 'minor',
      symbol = 'ABCDE',
      gene_id = 'SAMPLEGENE'
    )
  )

usethis::use_data(snp_annotations, overwrite = TRUE)