There are subtle differences in data available on the DepMap website and via the package. For example:

• At https://depmap.org/portal/gene/GPX4?tab=dependency&dependency=Chronos_Combined page, select "Show lineage subtypes" and see under "Soft Tissue" barplot "Rhabdomyosarcoma (13)" - 13 cells are expected.

• In the data obtained as

  library(depmap)
  crispr_data <- depmap_crispr()
  gpx4_data <- subset(crispr_data, gene_name == "GPX4")
  metadata <- depmap_metadata()
  gpx4_merged <- merge(gpx4_data, metadata, by = "depmap_id")
  

  Sorting by "lineage", then by "lineage_subtype" columns yields 12 rhabdomyosarcomas in the "lineage_subtype" column.

This is just one reproducible example. There are other subtle differences. Why it may be?