NEWS

---

dbparser 2.3.0

NEW FEATURES

• New Parser: parseHGNC()
  • Parses the HUGO Gene Nomenclature Committee (HGNC) complete set (hgnc_complete_set.txt) and, optionally, its withdrawn entries (withdrawn.txt).
  • Returns a dvobject with gene, gene_name (every approved, alias and previous symbol and name, one row each) and withdrawn (one row per withdrawn ID and the ID it was merged into or split to) tables.
• New Parser: parseHealthCanadaDPD()
  • Parses Health Canada's Drug Product Database (DPD) data extract: the marketed, approved, cancelled and dormant product sets, from their zips or unzipped folders.
  • Returns a dvobject with all 13 tables of Health Canada's read-me, with the read-me's column names, typed codes and dates, and an extract column naming each row's product set. Statuses are kept as written; the parser does no status grouping.
• New Parser: parseOpenFDA()
  • Parses openFDA's Drugs@FDA and NDC directory downloads, zipped or unzipped, without extracting them to disk.
  • Returns a dvobject with one table per record level (applications, products, active ingredients, submissions, submission property types, application documents and the openfda annotation for Drugs@FDA; products, active ingredients, routes, pharmacologic classes, packaging and the openfda annotation for the NDC directory), field names as in the files. Every record is kept as written, including tentative approvals; the parser does no status grouping.
  • Adds jsonlite to Imports.
• Source metadata on every dvobject
  • Every parser now records where its data came from in one attribute, source_metadata: release, release date, download date, source URL, licence (SPDX id), licence URL, the attribution or citation the source asks for, the parsed files with their SHA-256 hashes, and the dbparser version. Unknown values are NA, never guessed.
  • Merge and subset functions keep the record of every source.
  • File hashes need R 4.5.0 or later (tools::sha256sum()); on older R they are NA.
• New Utility Function: get_source_metadata()
  • Returns a dvobject's source metadata as a tibble, one row per source, with the files as a list column.
• New parsers name their version arguments release and release_date; existing parsers keep db_version and db_exported_date for now.
• New Parser: parseHPO()
  • Parses the Human Phenotype Ontology (HPO) gene-phenotype associations distributed by Harmonizome (gene_attribute_edges.txt, plain or gzip-compressed).
  • Returns a dvobject with associations, genes and phenotypes tables. Optional minimum_weight, db_version and db_exported_date arguments filter associations and record dataset metadata.
• New Integration Function: merge_drugbank_hpo()
  • Links DrugBank target genes to HPO phenotypes through gene symbols and adds drugbank_id and drug_name to each gene-phenotype association.
  • Chainable with merge_drugbank_onsides() and merge_drugbank_twosides().
• New HPO helpers: search_phenotypes() and get_genes_for_phenotype()
  • search_phenotypes() finds phenotypes by keyword and lists their genes.
  • get_genes_for_phenotype() returns the genes for an HPO id (e.g. "HP:0001250") or a phenotype name.

BUG FIXES

• parseDrugBank() extracts a zip input into a temporary folder that is removed after loading. Earlier versions extracted it next to the input file, which for the bundled example is the installed package. The message now reads "extracted".
• The merge functions no longer give tidyselect .data deprecation warnings.
• CETT actions and polypeptides$general_information tables now include a drugbank_id column (#210).
  • Earlier versions keyed these tables only by the carrier, enzyme, target or transporter id (e.g. target_id). Many drugs share a target, so an action could not be traced back to its drug, and identical actions from different drugs were collapsed into a single row.
  • Join CETT tables on both drugbank_id and the CETT id to get one row per drug–protein–action.

DOCUMENTATION

• New vignette vignette("getting_started") parses the bundled DrugBank record and HGNC files and shows their tables and source metadata. It replaces the DrugBank tutorial as the package's introduction on CRAN.
• The three vignettes with canvasXpress charts (the DrugBank tutorial, integrated pharmacovigilance and HPO target phenotypes) are now articles on the package website only, unchanged, at https://docs.ropensci.org/dbparser/articles/. This keeps the installed package small.
• New vignette vignette("cett_mechanisms") shows how to use the drug link in CETT actions to find drugs that share a mechanism. It uses lepirudin and bivalirudin, both direct thrombin inhibitors, as the example.
• New article Anticipating On-Target Drug Effects with HPO Gene-Phenotype Data links DrugBank targets to HPO Mendelian phenotypes to anticipate on-target drug effects for bivalirudin, leuprolide and calcitriol, with canvasXpress plots.
• show_dvobject_metadata() has a working example.
• The README has two new sections: "Commercial support and extensions" and "Used in published research".

COMING NEXT

• Parsers for France's public medicines database (BDPM), openFDA drug labels and FAERS summaries, and African pharmacogenomics (PGx) summary tables.

dbparser 2.2.1 (2026-01-08)

This is a major feature release focused on expanding dbparser's capabilities into real-world pharmacovigilance and drug-drug interaction analysis. The integration engine has been formalized around a "Hub and Spoke" model, with DrugBank acting as the central hub.

NEW FEATURES

• New Parser: parseOnSIDES()

  • Parses the relational CSV files from the OnSIDES database, a modern resource for adverse drug events extracted from FDA labels.
  • Returns a dvobject containing the 7 core relational tables and an optional high_confidence summary table.

• New Parser: parseTWOSIDES()

  • Parses the TWOSIDES database, the leading resource for drug-drug interaction (DDI) adverse event signals from real-world data.
  • Returns a dvobject containing the drug_drug_interactions table.
  • The parser correctly handles known column name misspellings in the source data (e.g., drug_1_rxnorn_id).

• New Integration Function: merge_drugbank_onsides()

  • Merges a DrugBank dvobject with an OnSIDES dvobject.
  • Automatically creates an enriched integrated_data list, linking OnSIDES data to DrugBank IDs via RxNorm CUIs.
  • The function is chainable, meaning it can be used in a %>% pipeline after other merge functions.

• New Integration Function: merge_drugbank_twosides()

  • Merges a DrugBank dvobject with a TWOSIDES dvobject.
  • Performs a "double join" to enrich both drugs in an interaction pair with their DrugBank IDs and names.
  • Uses a robust "union" logic to keep interactions even if only one of the two drugs is present in the input DrugBank object.
  • The function is also chainable.

• Subset a DrugBank dvobject function: subset_drugbank_dvobject()

  • Intelligently filters a DrugBank dvobject to retain only the data associated with a specified list of drugbank_ids. It correctly handles the deep, multi-level nested structure of the entire object, including the complex relationships within the cett list.

• Subset an OnSIDES dvobject function: subset_onsides_dvobject()

  • Intelligently filters an OnSIDES dvobject by cascading filters through the relational tables, ensuring the final subset is self-consistent.

• Adding metadata to existing dvobject objects using function: add_database_info()

• Major enhancements to function show_dvobject_metadata()

  • Displays information about passed dbobject object including basic info, database metadata, and all data.frames contained within nested lists.

• Run vignette("dbparser_2_2", package = "dbparser") for more info

DOCUMENTATION

• New Vignette: "Integrated Pharmacovigilance"
  • A comprehensive new tutorial demonstrating a full, three-way integration of DrugBank, OnSIDES, and TWOSIDES.
  • Includes a complete scientific case study analyzing single-drug vs. polypharmacy risks.
  • Introduces a reproducible example workflow using a pre-computed RDS data object hosted externally to keep the package lightweight.
  • Run vignette("drugbank_nside", package = "dbparser") for more info
• Updated existing vignette and package Readme with enhanced examples

BUG FIXES & MINOR IMPROVEMENTS

• Several minor fixes are done

---

dbparser 2.0.3 (2024-04-20)

Enhancements

• Updated unit tests with latest DrugBank DB info v 5.1.12

Bugs Fixes

• Fixed "Drug Targets has two drugbank_id columns" bug (#163)

---

dbparser 2.0.2 (2024-02-16)

Enhancements

• Replaced "-" with "_" in tibbles column names (#111)
• Renamed "primary_key" to "drugbank_id" in parsed drugs "general_information" tibble (#111)
• Replaced "parent_key" keywords in different tibbles column names with original parent name (#111)
• Updated Vignette to use interactive canvasXpress package plots

---

dbparser 2.0.1 (2023-03-27)

Bugs Fixes

• Fixed output references (#147)
• Updated parseDrugBanKparameters default values (#146)
• Fixed package documentation references (#144)
• Fixed CRAN error on some of linux info (#145)

---

dbparser 2.0.0 (2023-03-17)

Breaking changes

• Deprecated saving parsed data into given database (#140)
• Deprecated saving parsed data into a csv file (#140)
• Deprecated old structure public methods (#141)
• Updated minimum R required version to 3.5 (#143)

New features

• Introduced new methods for paring DrugBank DB and returning dvobject (#141)

Major Updates

• Updated unit tests to work with new data structure (#141)

Minor Fixes

• Removed RMariaDB dependency (#129)
• Fix pkgdown configuration for reference (#136)

---

dbparser 1.2.0 (2020-08-26)

UI Changes

• Introduce progress bar in parser functions

New Parsers

Collective Parsers

• drugs, cett and References Parsers

Elements Parsers

• attachments parsers for drugs and CETT
• drug_pharmacology parser
• Rename drugs_books parser to drugs_textbooks
• Rename drug_all parser to run_all_parsers
• Rename drug parser to drug_general_information

Documentation Update:

• Add returned parsed data structure
• Explain the returned data functionality as a whole and for each elements
• Point out to related/similar parsers

Package design

For those who thinking to contribute in dbparser, now parsers are implemented as R6 classes.

Minor Fixes

• Update database saver functions to accommodate new DrugBank data size.

---

dbparser 1.1.2 (2020-06-08)

Major Changes

• Enhance many memory and performance issues for many parsers.
• Change the drug classification representations to extract more useful information.

Minor Changes

• Change some drug tibbles features names

DEFUNCT

• Size columns in drugs main table is no longer exist, will do full statistical analysis later using dvminer package.

---

dbparser 1.1.1 (2020-05-10)

• Fix column size issue while importing into SQL Server (#91)
• Fix dbparser and upcoming CRAN release of dplyr issues (#92)
• Fix CRAN Notes (#93)
• Fix package documentation and site references

---

dbparser 1.1.0 (2020-02-21)

Major Changes

• Functions have been splitted into 6 categories DrugBank Database Loading, Carriers, Targets, Transporters, Drug and common parsers. All function names are changed to reflect the function family. The related documentation is also updated (#66, #75).
• dbparser now can cite the package by calling citation("dbparser") (#71).
• Adding more user friendly error messages (#76, #81).
• User can now pass DBI database connection to parser functions as an argument beside SQLite and MariaDB (#87).

DEFUNCT

• open_db, open_mdb and close_db functions are no longer supported. Creating and maintaining database is completely user responsibility and the database connection can be passed to parser functions (#87).

DOCUMENTATION FIXES

• New tutorials for how to use dbparser have been created (#78, #79).
• Contribution guide has been added.
• Code of conduct has been added (#70).
• Enhance function reference documentation to include section for each type (#68).

---

dbparser 1.0.4 (2019-08-28)

• Fix save drugs tibbles as csv several issues.
• Update SQL database tibbles saver functions.
• Update SQL database saver functions documentations.
• Support MariaDB and introduce related functionalities.

---

dbparser 1.0.3 (2019-07-11)

• Fix CRAN errors and notes

---

dbparser 1.0.2 (2019-07-08)

• Fix zip file location issue
• Replace Secondary and third keys columns from drug framework with other_keys column that contains any other keys that might exist in addition to the primary key
• Add average-mass, monoisotopic-mass and calculated-properties parsers.
• Support saving parsed drugs related parsed database as csv

---

dbparser 1.0.1 (2019-04-16)

• Fix CRAN Note
• Improve documentation
• Refactor unused functions
• Remove Count features from drug data set
• Fix several typos in documentation and code
• Fix consistency issue of CLASS of tibbles Returned by dbparser
• Check if DrugBank database exist before parsing
• Add support for international_brands and salts elements
• Properly rename some features to have clear names
• Reduce datasets size by getting unique rows only
• Support reading zip file containing DrugBank xml database

---

dbparser 1.0.0 (2018-12-17)

• Initial release that contains core functionalities