Get information about human genes from RefSeq

get-human-gene-info.R

library(tidyverse)

# Get Gene Summary info
gs_orig <- read_tsv("https://ftp.ncbi.nlm.nih.gov/gene/DATA/gene_summary.gz")
gs <- gs_orig |>
  janitor::clean_names() |>
  set_names(str_replace, "number_tax_id", "tax_id") |>
  filter(tax_id==9606) |>
  distinct()
gs

# Get gene information for human
gi_orig <- read_tsv("https://ftp.ncbi.nlm.nih.gov/gene/DATA/GENE_INFO/Mammalia/Homo_sapiens.gene_info.gz")
gi <- gi_orig |>
  janitor::clean_names() |>
  set_names(str_replace, "number_tax_id", "tax_id") |>
  filter(tax_id==9606) |>
  distinct()
gi

# Get gene to gene ontology
go_orig <- read_tsv("https://ftp.ncbi.nlm.nih.gov/gene/DATA/gene2go.gz")
go <-
  go_orig |>
  janitor::clean_names() |>
  set_names(str_replace, "number_tax_id", "tax_id") |>
  filter(tax_id==9606) |>
  distinct() |>
  filter(category=="Process") |>
  summarize(biological_process = paste(go_term, collapse="; "), .by="gene_id")
go

# join them all together
g <-
  gi |>
  filter(type_of_gene=="protein-coding") |>
  inner_join(gs, by="gene_id") |>
  left_join(go, by="gene_id") |>
  select(symbol, description, gene_info=summary, biological_process) |>
  distinct()

# Write out files
g |> write_csv("hs_gene_info.csv")
g |> nanoparquet::write_parquet("hs_gene_info.parquet")