Skip to contents

Changelog

DO.utils (development version)

General

Updated

  • is_invariant() no works for more than just character & numeric vectors, with new list and data.frame methods and a default method that should be able to handle more cases (and replaces the character method).

DO Management & Analysis

Updated

  • read_omim() now additionally parses official API-key requiring phenotypicSeries.txt downloads and may be able to handle additional API-key requiring downloads.
  • plot_citedby(): * _[BREAKING CHANGE]_color_setargument now requires names and one color for each of the 7 possible publication types when specifying colors manually. *retracted` argument added to specify how retracted articles should be managed.

New

  • download_omim() downloads official API-key requiring files directly from OMIM (e.g. mim2gene.txt, phenotypicSeries.txt, etc.).
  • extract_ordo_mappings() extracts mappings from Orphanet Rare Disease Ontology, in native format as oboInOwl:hasDbXref with Orphanet’s text-based predicate modifiers, or as SKOS (supplemented with filler doid: predicates where SKOS predicates don’t exist.

DO.utils 0.3.1

Dependency Update

  • dplyr > v1.1.0 now required.

General

New

  • elucidate() (generic) describes the data in a given object. Currently, only omim_inventory has a defined method.

DO Management & Analysis

Updated

  • read_omim() now additionally parses official OMIM downloads of search results and phenotypic series titles.
    • Includes omim_official attribute to indicate if the source was an official download.
    • If input is an official source, the output class will indicate the type.
    • keep_mim arg can be used to filter OMIM search results.
  • tidy_sparql() now removes ? from column names and has the new argument lgl_NA_false for specifying whether NA values should be replaced with FALSE in logical columns.
  • write_gs.omim_inventory() now has a datestamp method.

DO.utils 0.3.0

General

Updated

  • Aligned read_delim_auto() more closely with readr::read_delim() so it could handle compressed input.
  • Broadened unique_if_invariant() no longer uses it’s own methods and instead relies on base::unique(). This may have some unintended consequences, particularly where custom methods of unique() are defined but it works for more inputs, better matching expectations.
  • format_html() txt argument has been renamed to text to align it more fully with expectations.

New

DO Management & Analysis

WEBSITE-Supporting Updates

  • Fix make_use_case_html() to have case-insensitive sorting.
  • plot_citedby()
    • Now accepts manually-defined color sets, in addition to color sets provided by DO.utils.
    • Default plot size changed to better fit new position on disease-ontology.org statistics page.

Other Updates

  • robot()
    • .path argument renamed to .robot_path to avoid use in other functions without changing the name.
    • Now informs when testing and caching a ROBOT executable for future use.
  • onto_missing() was poorly designed and has been deprecated. To determine which OMIM entries are present in the DO as mappings, use inventory_omim() instead.

New

  • NEW FUNCTIONALITY to speed up curation of OMIM mappings!!!
    • read_omim() reads data copied or downloaded from OMIM.
      • Previously an internal function with limited capability to handle specific copy/paste operation from OMIM.
      • Now expanded to read data downloaded from omim.org phenotypic series pages using the “Download as” button and to handle all copy paste of tabular data from omim.org without the need for manual corrections.
      • No longer returns tidy_label and provisional columns, as these were not particularly useful, and instead includes omim and geno_inheritance columns to help with curation.
    • inventory_omim() compares OMIM entry records against mappings in the DO and reports whether they exist, with accompanying DO class data when they do.
    • write_gs() (generic) writes data from DO.utils created classes to Google Sheets.
      • omim_inventory is the first method.
  • read_ga() to read Google Analytics data exported to .csv file.
    • Eliminates the need for time-consuming, corrections to get a GA exported file into a tidy format for further use
    • Can optionally read multiple tables from a single file (most exports have two).
    • Can NOT merge GA data that has been split over multiple files due to size. These must be merged manually but this should be trivial.

DO.utils 0.2.10

DO Management & Analysis

  • onto_missing() & tidy_sparql() output now has improved column types (no longer all character vectors).
  • append_to_url() has a new named URL option, “DO_website”, for direct link to disease info on disease-ontology.org.

Assessing Resource Use

  • citedby_scopus() will no longer retain responses with zero results and gained a new argument no_results to control how these are signaled to the user, making it more consistent with citedby_pubmed().

DO.utils 0.2.9

DO.utils 0.2.8

Highlights

  • Adds onto_missing() and character length-sorting functions.
  • More support for creating links from CURIEs.
  • Includes fixes to eliminate warnings from use of tidyverse (#15) and errors due to updates in some tidyverse packages.

General Utilities

New

  • length_sort(): Sorts vector elements by character length.
  • length_order(): Sorts data.frames by character length of elements in specified column(s).
  • iff_all_vals(): Tests if all values are present in a vector and ONLY those values are present.

Updated

  • drop_blank(): Now a generic with character and list methods.
  • vctr_to_string(): Now always returns NA when only input is NA, even when na.rm = FALSE; previously returned "NA".

DO Management & Analysis

New

  • is_curie(): Tests for CURIEs in character vectors, according to a specified definition that always conforms to W3C CURIE Syntax 1.0.
  • onto_missing(): Compares tsv/csv data with data in the ontology to identify data that may be missing. Optionally returns data that is present.

Updated

  • robot() errors are now signaled in R and no longer specifies a max heap size when using a robot.jar file.
  • to_curie()/to_uri() now appropriately remove brackets from URIs and handle delimited input.
  • tidy_sparql() has the new as_curies argument and converts IRIs to CURIEs by default.
  • ns_prefix now includes more namespace prefixes, including those for MeSH and UniProt.
  • append_to_url() is now vectorized and can append to additional prefixes, including anything in ns_prefix and URLs commonly used on disease-ontology.org for cross-references.
  • format_url() no longer uses NA in the txt argument as text input.
  • build_hyperlink() takes advantage of the updates to append_to_url() and format_url().

DO.utils 0.2.7

Dependencies

  • reticulate updated to >=v1.28 in an effort to resolve python package installation issues; see https://github.com/DiseaseOntology/DO.utils/issues/12.
  • stringr updated to >= 1.5.0, for access to new str_escape() function.
  • igraph added for new tidygraph/graphml functions. This change doesn’t affect much since tidygraph is already a dependency that itself depends on igraph.
  • Now SUGGESTS the keyring package for API key management.

Announcements

  • Migrated DO.utils repository to the DiseaseOntology organization.
  • DO.utils documentation is now available on the web at https://diseaseontology.github.io/DO.utils/ with significant updates supporting citation-based assessment of use workflow.

DO Management & Analysis

New

  • extract_as_tidygraph(): Extracts nodes and relationships identified in a SPARQL query from RDF/XML file and returns it as a tidygraph.
  • write_graphml(): Writes a graph object (tidygraph/iGraph) to a .graphml file.

Updated

  • [BREAKING CHANGE] robot() wrapper function updated to make it easier to use when programming.
  • plot_branch_counts() now:
    1. Identifies the count of classes in each branch that are asserted or inferred.
    2. Uses data directly from a local copy of the HumanDiseaseOntology repo instead of manually copied release notes.
    3. Gained the aspect_ratio argument.
  • plot_xref_counts() now uses data directly from a local copy of the HumanDiseaseOntology repo instead of manually copied release notes.
  • plot_citedby() gained the color_set argument to permit more flexible color choice.
  • DO_pubs updated with 2023 complex disease paper.
    • NOTE: The 2023 Database paper describing the ‘Assessing Resource Use’ workflow was NOT added because it is not a publication describing use of the DO).
  • DO_colors now include accent colors generated as part of the DO-KB addition to the website. These colors are available in standard, “_mid”, and “_light” versions.

Assessing Resource Use

New

  • ‘Assessing Resource Use: Obtaining Use Records’ tutorial/vignette added. Describes how to set up DO.utils and execute functions to support the ‘Assessing Resource Use’ workflow.
  • tidy_pub_records(): creates a tibble with more limited information from Scopus and PubMed references; includes only the columns: first_author, title, journal, pub_date, doi, pmid, scopus_eid, pub_type, and added_dt.
  • set_scopus_keys(): makes Scopus API key and/or insttoken available for use during an R session.
  • set_entrez_key(): makes Entrez Utils API key available for use during an R session; imported from rentrez package.

Updated

General Utilities

Updated

  • to_range() now returns NA when passed empty vectors.
  • citedby_scopus() has a new insttoken argument.
  • collapse_col() gained all the methods of collapse_col_flex(), along with na.rm argument that can be used by all methods.
  • append_to_url() and build_hyperlink() no longer add a trailing slash to the end of URLs when there is not one. Also, a new sep argument has been added to provide greater control.
  • [BREAKING CHANGE] format_hyperlink() preserve_NA argument removed and replaced with preserve argument. With this change, the output value when a URL is missing will be either the URL (i.e. NA) or the text passed to txt. This allows more flexibility in the output to support more use cases.
  • format_hyperlink() now warns when values are passed to ... when as does not equal “html” to reduce the likelihood of losing arguments silently.

DO.utils 0.2.6

Change license to CC0 1.0 Universal to match standard for the Human Disease Ontology project and in preparation for use in resource use assessment publication.

DO.utils 0.2.5

Deprecations

Website * make_user_list_html() is deprecated because the user/use case information on disease-ontology.org was moved from the ‘Collaborators’ page to the new ‘Use Cases’ page. Replaced by make_use_case_html().

Breaking Changes

Website * plot_*() no longer include the datestamp in saved file names.

Formatters * format_doid() parameters changed: * allow_bare renamed to convert_bare * validate_input added to allow invalid input to pass-through without modification.

Cited by / Search * read_pubmed_txt() now parses IDs (PMID, PMCID, & DOI) from citations and returns a data.frame that includes a record number, IDs, and full citations instead of a vector of citations.

Bug Fixes

Cited by / Search * extract_pmid() updated to recognize 1- to 8-long PubMed IDs which should cover the whole set of actual PMIDs; previously limited to recognizing 8-long PMIDs. * as_tibble.esummary_list() fixed error due to reduced data output (fewer columns of information) from pubmed_summary() caused by API changes. * match_citations() now matches DOIs in a case insensitive manner bringing it into compliance with the DOI spec.

Updates

URLs * append_to_url(): * Gained a new parameter preserve_NA, which allows NA values to pass through instead of being appended to the URL. * Added ‘github’ and ‘orcid’ as named URLs that might be appended to (via get_url()).

Datasets * More prefixes in ns_prefix and new prefix subsets added: * not_obo_prefix: Subset of ns_prefix with everything except OBO ontology prefixes (e.g. ‘dc’, ‘terms’, ‘skos’, ‘owl’, etc.). * obo_prefix: Subset of ns_prefix with standard OBO ontology prefixes and namespaces. * obo_prop_prefix: New set of prefixes created to represent frequently used OBO ontology property prefixes. NOTE: There is one per ontology but these may not exist or may not be the actual property prefix used by the stated ontology… use with caution.

General Utilities * unique_to_string()/vctr_to_string(): added sort and other arguments for control of sorting.

New

Ontology Extracters/Modifiers * extract_*_axiom() family: Extract equivalentClass (‘eq’), subClassOf (‘subclass’), or both (‘class’) logical axioms. * queue_xref_split(): Creates a ‘curation queue’ of diseases that may need to split because they have multiple cross-references from the same source. * tidy_sparql(): Tidies SPARQL query results.

Website * update_website_count_tables(): Update counts in tables on ‘DO Imports’ and ‘DO Slims’ pages with data from a specified release of doid-merged.owl. Updates data in place. * make_use_case_html(): Produces the html for the new ‘Use Cases’ page, split into 3 files, 1 per section for: Ontologies, Resources, and Methodologies. * Does not update data in place. HTML for the rows & cells must be copied and pasted over the HTML for each section in the ‘Use Cases’ file. * Content is sorted alphabetically by column. * make_contributor_html(): Produces HTML list of contributors as <li> elements for disease-ontology.org, including links to Github and ORCID, as available.

URLs * format_hyperlink(): Converts URLs into hyperlinks for Google Sheets, Excel, or HTML. * build_hyperlink(): Shorthand for common append_to_url() plus format_hyperlink() combination.

Cited by / Search * pub_id_match (DATA): A named character vector of regex’s to identify/extract publication IDs (currently PMID, PMCID, DOI, & Scopus EID).

General Utilities * sandwich_text(): Pastes text around strings. * wrap_onscreen(): Wraps messages to be printed on the screen. * invert_sublists(): Swaps the list elements between depths 2 & 3, essentially inverting the grouping. * lengthen_col(): Splits column(s) by a delimiter and lengthens the data.frame so each value is in its own row. * NOTES: * This is the reverse of collapse_col() but will not recreate the original data.frame after round trip in most cases. * Uses unnest_cross() internally so the results will always be the cartesian product of lengthened columns. * count_delim(): Counts the values in delimited columns; essentially the combination of lengthen_col() and dplyr::count().

Type Predicates * is_valid_obo(): Tests whether the elements of a character vector are ‘valid’ OBO Foundry IDs (based on formatting, not actual existence).

Formatters * format_obo(): Formats OBO Foundry IDs. * format_axiom(): Formats OWL functional syntax EQ/SubClassOf axioms to be more human readable, similar to that of Protege.

DO.utils 0.2.4

New

Testing * is_boolean(): T/F type predicate. * write_access(): Test for file write existence with write access.

Data Conversion * to_curie() & to_uri(): Convert between URI & CURIEs. * to_range(): Convert vector of values to ranges (output as single string).

Data.frame Manipulation * append_empty_col(): Add empty columns to a data.frame. * unnest_cross(): Unnest list columns in data.frame always creating the cartesian product. * Useful for expanding list columns produced by some SPARQL queries.

Datasets * ST_pubs: Information about official publications describing the Symptom (SYMP) and/or Pathogen Transmission (TRANS) ontologies. * Currently, only one conatins one publication. * ns_prefix: Named character vector of common namespace-prefix pairs used in DO and/or OBO ontologies.

Updates

  • cast_to_string renamed to collapse_to_string()
  • DO_pubs now includes lens_id with lens.org identifiers for each publication.

Dependencies

  • Imports ‘tidyselect’, which is explicitly required for unnest_cross but is used throughout DO.utils to enable tidyverse-style semantics (via dplyr).

DO.utils 0.2.3

Cited By – Updates

  • citedby_pubmed(), equivalent to citedby_scopus(), now available (uses citedby_pmid() and pubmed_summary() internally.
  • match_citations() improvments:
    • BREAKING CHANGE – add_col argument changed from boolean to NULL (replaces FALSE) or the name of the column (replaces TRUE).
    • Bug fixes & message improvements.
  • extract_pmid() has a new argument, no_result, which provides control over the condition signaled when no results are found (error, warning, message, or none); previously a simple error was signaled. For any condition signaled, there is now an additional class no_result to improve condition handling.
    • For the elink method the default for no_result is still “error” while for elink_list the default is now a warning.
  • as_tibble.esummary_list() bug caused when some results have no data fixed – caused errors in as_tibble.esummary_list_nested() precipitated by the tidyverse’s move to more strict vector merging.

IMPORTANT NOTE: The “cited by” functionality of DO.utils may no longer be improved because recent review of Lens.org results suggests it may be a good replacement for this PubMed + Scopus search & merge strategy.

If improvements are made they will likely facilitate one or more of:

  • Merging PubMed and Scopus “cited by” results, probably using standardize() (or similar).
  • Reducing data requested from the APIs by implementing timeframe parameters in citedby_*().

DO.utils 0.2.2

  • Bug fix to correct error in format_subtree() when the subtree did not have any classes with multi-parentage. (Error label: “no fill needed”)

DO.utils 0.2.1

New

  • Create a text-based subtree/hierarchy.
  • Manage DOIDs.
    • is_valid_doid() tests whether inputs are valid DOIDs. Note that mutliple formats are considered valid.
    • format_doid() converts between valid DOID formats.

Updates

  • DOrepo() and owl_xml() no longer fail silently when a file/directory does not exist. pyDOID now verifies file paths on instantiation of the underlying objects.

Dependencies

DO.utils 0.2.0

Data

  • Added DO Nucleic Acids Research 2022 publication data to DO_pubs.

Dependencies

  • R packages:
    • reticulate >= v1.23 required.
    • tidyr v1.2.0 introduced breaking changes by requiring “safe” type casting (implemented via vctrs).
  • Python dependency: pyDOID

General Purpose

  • Created collapse_col_flex() to collapse data frame columns more flexibly.
    • Adds two new methods beyond “unique”: “first” & “last”.
    • Adds the ability to collapse columns using different methods.
  • Created wrapper functions for pyDOID classes.
    • DOrepo() wraps the pyDOID.repo.DOrepo class
    • owl_xml() wraps the pyDOID.owl.xml class

Graphics / Website

  • Updated:
    • plot_citedby() to a stacked bar chart showing publication types.
    • DO_colors to include saturated versions (names prefixed with sat_).
    • Functions generating html have been updated to match html style guide standards.
  • Created:
    • theme_DO() a ggplot2 plotting theme for DO.
    • plot_def_src() to display the number of times a source is used to support disease definitions in the ontology (designed for disease-ontology.org/about/statistics).

Cited by

URLs

  • Created:
    • Functions to read the doid-edit.owl file and extract URLs (+ helpers).
    • Functions designed for URL validation.
      • validate_url() + helpers
      • NOTE: Helpers for robots.txt respectful validation remain INCOMPLETE and care should be taken not to overwhelm web servers with requests.

DO.utils 0.1.7

  • Setup package to wrap python via reticulate package.
  • Added functionality to predict mappings using GILDA grounding (a type of lexical string matching/natural entity recognition) via python and the python modules pyobo, indra, and gilda. New functions:

DO.utils 0.1.6

General

  • Added DEPENDENCIES on ggplot2, googlesheets4, and glue.
  • Renamed match_citations_fz() to match_fz().
  • Added cast_to_string(), a more generalized version of vctr_to_string() that accepts multiple inputs (similar to paste()).
  • Added function to partition() vectors into groups with n elements per group.

Data

  • Added latest official DO publication to DO_pubs.
  • Added official DO_colors.

Feature: Website Updates

Internal Use Only

  • Added to_character(), helper for cast_to_string(), to reduce lists and data frames to character vectors while limiting data loss.
  • Added html_in_rows(), helper for make_user_list_html(), to format html elements in rows (with optional row & cell attributes).
  • Added Google sheets identifiers for programmatic access.

DO.utils 0.1.5

General

  • Added a NEWS.md file to track changes to the package.
  • Created/updated various helpers to manage data (get it, make it easier to save/track with version control):
    • download_file() to flexibly handle multiple downloads.
      • Created download_status Ref Class to manage downloads based on exit code/status.
    • confine_list() / release_list() to reversibly convert a list column to a character vector (using json).
    • is_invariant() to test for vectors with only 1 value; with methods for character & numeric vectors.
    • unique_to_string() to collapse vectors to strings.
    • unique_if_invariant() to conditionally collapse vectors with only 1 value.
    • Added na.rm argument to all vctr_to_scalar functions.
    • replace_*()
      • NA, method for lists.
      • NULL, to replace NULL values in lists recursively.
      • blank, to replace “” values.
    • collapse_col() to collapse 1 or more specified columns in a data frame by concatenating unique values together, while preserving unique values in all other columns.
  • Created download_obo_ontology() to download 1 or more ontologies maintained by the OBO Foundry.

Data

  • Added more to DO publication info.
  • Added OBO Foundry metadata.

Feature: Alliance

  • Made it possible to count Alliance terms for a subset of DOIDs.
  • Increased record type count options (arg: record_lvl) –> “full_record”, “disease-object”, “disease”, “object”

Feature: citedby