DrugBank_parse

Parse downloaded DrugBank XML files into drug, target, indication, and drug-target mapping tables.

This repository now contains two generations of code:

• dev/: the current package-first rewrite, with shared schema definitions and matching Python/R core parsers.
• v1, v2, vR: legacy scripts kept for reference and compatibility experiments.

The rewrite is the recommended path for new work.

Current Status

The dev/ implementation currently supports the core DrugBank output contract:

• Python package: dev/python
• R package: dev/R
• Shared schema: dev/schema
• Expected fixture CSVs: dev/fixtures/expected/core

Both Python and R implementations parse the bundled test-database.xml fixture and export the same core CSV files.

Core Output Tables

• drugs.csv
• targets.csv
• drug_target.csv
• drug_indication.csv
• target_drug_indication.csv

Python Quick Start

Run these commands from dev/python.

Install the package with test dependencies:

python -m pip install -e ".[test]"

Run tests:

python -m pytest -q

Parse the bundled fixture:

python -m drugbank_parse.cli --input ..\..\test-database.xml --profile core --outdir ..\tmp_core_output

Use the package API:

from drugbank_parse import parse_drugbank_xml, write_drugbank_tables

result = parse_drugbank_xml("drugbank_5-1-12.xml", profile="core")
write_drugbank_tables(result, "output")

R Quick Start

Run these commands from dev/R.

Run tests:

Rscript -e "testthat::test_local(reporter='summary')"

Use the package API after installing or loading the package:

library(drugbankparse)

result <- parse_drugbank_xml("../../test-database.xml", schema_dir = "../schema")
write_drugbank_tables(result, "tmp_r_core_output", schema_dir = "../schema")

Directory Guide

• dev/README.md: detailed notes for the current rewrite.
• dev/benchmarks: fixture and full-XML benchmark scripts.
• dev/python: Python core parser package and tests.
• dev/R: R core parser package and tests.
• dev/schema: shared YAML table/profile definitions.
• dev/fixtures: small fixture data and expected core CSV outputs.
• v1: early Python parser with faster parsing but harder-to-read code.
• v2: more readable Python parser with slower file processing.
• vR: older R scripts with low memory usage but much slower runtime.
• test-database.xml: small DrugBank XML fixture for tests.

DrugBank XML Input

Download full DrugBank XML releases from:

https://go.drugbank.com/releases/latest

The legacy scripts were tested against DrugBank release v5.1.12. The current rewrite uses test-database.xml and shared fixture CSVs as its regression contract while the full-file parser is being expanded.

Notes

The rewrite is intentionally schema-driven so Python and R exports stay aligned. Add new fields or tables by updating dev/schema first, then extending the Python/R parsers and expected fixtures together.