data-raw/metadata_tables.R

## code to prepare `metadata_tables` dataset goes here

library(org.Hs.eg.db)
library(tidyverse)

rename_columns <- function(input_names) {
  if ("Perturbagen" %in% input_names) {
    col_names <- c("SourceSignature", "Source", "SourceConcentration",
                   "SourceCellLine", "SourceTime")
  } else {
    col_names <- c("SourceSignature", "Source",
                   "SourceCellLine", "SourceTime")
  }

  col_names
}

oe_metadata <- read_tsv("raw/LINCS-OE-Metadata.xls") %>%
  select(-CGSID, -is_exemplar) %>%
  rename_with(rename_columns)

kd_metadata <- read_tsv("raw/LINCS-KD-Metadata.xls") %>%
  select(-CGSID, -is_exemplar) %>%
  rename_with(rename_columns)

cp_metadata <- read_tsv("raw/LINCS-Perturbagen-Metadata.xls") %>%
  select(-GeneTargets, -is_exemplar) %>%
  rename_with(rename_columns)

l1000_list <- read_tsv("raw/l1000_genes.tsv")

l1000 <- l1000_list %>%
  pull(HGNC) %>%
  unique %>%
  AnnotationDbi::select(org.Hs.eg.db, keys = ., columns = c("ENTREZID", "ALIAS", "SYMBOL"), keytype = "SYMBOL") %>%
  inner_join(l1000_list, by = c(SYMBOL = "HGNC")) %>%
  select(ENTREZID, L1000, SYMBOL, ALIAS)

usethis::use_data(oe_metadata, kd_metadata, cp_metadata, l1000,
                  internal = TRUE, overwrite = TRUE, compress = "bzip2", version = 2)
AliSajid/DrugDiscoveryPanCan documentation built on Feb. 4, 2021, 6:31 a.m.