Parsing PREDICT indications

library(dplyr)
library(DT)

options(stringsAsFactors = FALSE)

Indications from PREDICT

# map.df from msb201126-s4.xls
map.df <- file.path('data', 'omim-to-umls.tsv') %>%
  read.delim(check.names=FALSE) %>%
  dplyr::transmute(omim_id = `OMIM ID`,
                   umls_cui = `UMLS concept ID`,
                   umls_name = `UMLS concept name`)

# indication.df from a processed version msb201126-s1.xls
indication.df <- file.path('data', 'indications.tsv') %>%
  read.delim()

# add umls_ids to indications
indication.df <- indication.df %>%
  dplyr::left_join(map.df) %>%
  dplyr::select(drugbank_id, drugbank_name, umls_cui, umls_name, omim_id, omim_name) %>%
  dplyr::distinct(drugbank_id, umls_cui, omim_id)

## Joining by: "omim_id"

DT::datatable(indication.df)

indication.df %>%
  write.table(file.path('data', 'indications-umls.tsv'),
              sep='\t', row.names=FALSE, quote=FALSE)