oglcnac

oglcnac helps curate O-GlcNAcAtlas data files before they are published on oglcnac.org. The package keeps the workflow simple: work with CSV, TSV, or Excel files, validate the required Atlas columns, enrich UniProt fields, and export clean CSV files for the static website.

Main Workflow

library(oglcnac)

atlas <- read.csv("Atlas_unambiguous.csv")

validation <- validate_atlas_data(atlas, dataset = "unambiguous")
validation$valid

atlas <- process_tibble_uniprot_cached(
  atlas,
  cache_path = "~/.cache/oglcnac/uniprot-cache.rds"
)

export_atlas_csv(
  atlas,
  "atlas-records-unambiguous.csv",
  dataset = "unambiguous"
)

Use dataset = "unambiguous" for unambiguous sites, also called dataset-I. Use dataset = "ambiguous" for ambiguous sites, also called dataset-II. The package writes this value to the ambiguous column used by the public website, so the two Atlas datasets are not mixed.

Shiny App

oglcnac::launch_app()

The app supports:

• CSV, TSV, and Excel upload
• Atlas dataset selection
• Atlas schema validation
• cached UniProt enrichment
• processed CSV download

Website Export

The website source can generate static JSON directly from CSV files:

python3 frontend/scripts/generate_static_data.py \
  --atlas-unambiguous-csv atlas-records-unambiguous.csv \
  --atlas-ambiguous-csv atlas-records-ambiguous.csv \
  --ogt-pin-csv ogt-pin-records.csv

SQLite is still useful for legacy recovery, but CSV files are easier to review and should be the normal update format.