STRING network and interaction-partner functions now support directed regulatory networks and typed functional networks, with controls for physical and regulatory overlays. Network images can also hide regulatory signs while preserving edge direction.
rba_string_enrichment() gains enrichment scores, filtering, and term
merging by gene-set overlap. rba_string_enrichment_image() gains
corresponding filters and term merging, independent axis and sorting
controls, bar plots, GWAS and Hallmark categories, and the red_blue palette.
rba_metadata() retrieves optional request metadata attached to rbioapi
results. Metadata can be enabled for one call with metadata = TRUE or for
all later calls with rba_options(metadata = TRUE). It records the rbioapi
version and, for each request, its time, call, original httr response, and
the exact function(s) used to parse that response. Multi-request functions,
or multiple calls due to retries preserve request order. Metadata remains off
by default because it can substantially increase the size of returned
R objects.
rba_pages() now supports inclusive "pages:start:end" ranges in either
direction and arbitrary ordered page sets through page_arg and pages.
It accepts at most 100 unique positive whole-number pages and waits at least
two seconds between requests. Results preserve the requested order, retain
NULL pages, and use names such as page_1. The skip_error,
progress, and verbose arguments now provide explicit control over the
complete operation.
rba_reactome_search() searches Reactome by text with filters for species,
entry type, compartment, keyword, and scope, together with grouping and page
controls.
rba_enrichr_gene_sets() retrieves all gene sets, or one exact term, from
an organism-specific Enrichr library and can save the original GMT file.
rba_panther_genome() retrieves one page of genes from a PANTHER genome
and reports the current page, total pages, and number of genes in the genome.
rba_string_functional_terms() searches STRING functional terms by ID or
descriptive text for a species.
rba_uniprot_features_type() searches UniProt feature descriptions for a
chosen feature type, with optional annotation filters.
rba_uniprot_variation_locations() retrieves UniProt variants for paired
accession and sequence-position values. Results are returned as a list named
by accession or saved as PEFF.
rba_panther_enrich() gains request_mapped_genes and can return genes
mapped from the input, genes mapped from the reference set, or no mapped
genes. Input-gene mappings are returned by default.
rba_string_interactions_network() and rba_string_network_image() can
request a network with network_term_id instead of a list of IDs.
rba_string_annotations() gains only_pubmed; when TRUE, it takes
priority over allow_pubmed.
rba_reactome_event_hierarchy() gains pathway-only output and filters
based on an analysis token. rba_reactome_query() gains controls for
relationships, reference-entity summaries, and diseases.
rba_reactome_xref() now accepts several IDs and supports database filters
and result pages.
rba_mieaa_cats() gains a mode argument for all, default, or expert
categories. rba_mieaa_enrich() gains poll_interval and
poll_timeout to control how often a job is checked and how long polling
continues.
STRING requests now use the stable 12.5 address. Ordinary functional and
physical network schemas and default enrichment columns are unchanged,
but updated database evidence and annotations can change scores and results.
Enrichment figures now enforce the documented limit of 100 displayed terms.
Empty PNG enrichment figures, including high-resolution images, now return
a zero-length numeric image array instead of NULL.
rbioapi now rejects NA and NaN unless a function explicitly permits
them. Numeric identifiers, counts, sequence positions, page numbers, and
retry counts must be finite whole numbers where applicable. Invalid values
passed to rba_options() now produce an error instead of silently restoring
package defaults.
rba_mieaa_enrich_submit() no longer assumes human input and now requires
species as a supported abbreviation, scientific name, or taxonomy ID.
rba_panther_info(what = "families") now names its family result
family, correcting the previous familiy spelling.
rba_reactome_pathways_events() now returns complete contained-event
records as a named list rather than a data frame. Requests for one event
attribute preserve empty and structured values; if the response cannot be
separated reliably by event, the original response is returned with a
warning.
rba_string_map_ids() now uses STRING v12's get_string_ids service and
returns at most one match for each input ID. echo_query now defaults to
TRUE, queryIndex is returned as an integer, and saved responses use
.tsv. The limit argument is deprecated and has no effect.
rba_string_network_image() and rba_string_enrichment_image() now
return PNG results as decoded image arrays and SVG results as raw content.
rba_uniprot_rna_edit_search() renames variantlocation to
variant_location to match UniProt.
Enrichr functions now use the correct export and Speedrichr view web
addresses, accept positive whole-number list IDs, and use .tsv files for
standard enrichment and .json files for background enrichment.
Custom-background results preserve one row per enriched term instead of
creating extra rows from the returned gene list (#20).
JASPAR functions now default to and accept the 2026 release. Requests for collections, matrices, releases, sites, species, taxon groups, and TFFMs use current query values and supported taxon groups. The TFFM search alias, saved file names, and JASPAR error details are also corrected.
miEAA functions now use the current /mieaa/ web addresses and check
category names before submitting a job. Conversion results use .txt or
.tsv files to match their content. Over-representation analysis (ORA),
gene set enrichment analysis (GSEA), empty results, and numeric sorting are
handled correctly. Failed, malformed, or timed-out jobs stop promptly while
still respecting skip_error.
STRING requests now follow the current routes for ID mapping, interactions,
homology, enrichment, annotations, and images. Set-based operations remove
duplicate IDs and backgrounds, while ID mapping preserves duplicates and
input order. Functions subject to STRING's species limit now require
species for more than 10 unique IDs. Empty enrichment results and image
file names are also handled correctly.
Best-hit homology now correctly handles multiple target species.
Enrichment-figure grouping now accepts all documented tenths from 0.1 to 1.
PANTHER requests now follow the current service limits. Mapping accepts up to 5,000 IDs, enrichment checks the 100,000-ID limit, and TreeGrafter accepts sequences up to 50,000 characters. Paralog requests, family page counts, out-of-range pages, mapped-gene results, and temporary-file cleanup are corrected. Empty, single-record, and multi-record table-like results now have consistent data-frame forms while preserving nested fields.
Reactome analysis functions now use the current submission, mapping, report,
and download routes. Scalar IDs, vectors, tables, local files, and HTTP(S)
URLs are handled consistently, and local files use Reactome's current form
upload. Direct, token, and species analysis results now represent
pathways as a data frame, including a zero-row data frame when no pathways
are found.
Reactome content functions now use the current hierarchy, enhanced-query,
interaction, orthology, cross-reference, and exporter resources.
rba_reactome_analysis_token() makes species optional and defaults
p_value to 1. rba_reactome_exporter_diagram() now defaults
ehld to TRUE, matching Reactome's current behavior.
UniProt protein, feature, variation, proteomics, proteome, gene-centric, coordinate, taxonomy, and UniParc functions now use current web addresses, argument names, filters, page handling, and returned data. Search results are named by accession more reliably, and deprecated proteomics functions call their supported replacements.
rbioapi now detects errors reported in a response body with HTTP 200 and
applies skip_error in the same way as for regular HTTP errors.
Error messages from Enrichr, JASPAR, miEAA, PANTHER, Reactome, STRING, and UniProt now include the HTTP status and useful server details. Unrecognized response formats return the original text instead of hiding it. UniProt server errors are handled, and 4xx and 5xx responses are correctly described as client and server errors.
Response saving now more robustly supports Windows and Unix-like file paths containing backslashes, spaces, or quotes. Missing targets are handled consistently, and invalid paths produce clearer errors.
Required-argument detection, optional NULL handling, retry validation, and
request-header construction now behave consistently and report invalid
inputs more clearly.
Function manuals, argument and return descriptions, examples, vignettes, external links, and service citations are updated and standardized. Slow examples use smaller requests so checks and documentation builds take less time.
png moves from Suggests to Imports because STRING image functions
use it while the package is running. The minimum suggested testthat
version is now 3.1.7. Tests cover the new input checks, service errors,
metadata, multiple pages, and file saving, and the GitHub Actions files use
current templates.
Documentation is regenerated with roxygen2 8.0.0. The pkgdown reference lists the new functions, and README badges, links, examples, and development notices are updated.
Improved internal stability of Enrichr functions
Minor improvements and fixes.
Update functions to the latest corresponding API endpoints:
Added STRING function:
rba_string_enrichment_image()
Updated Enrichr function:
Added option to supply background genes
Updated PANTHER function:
added option to supply data frame to the enrichment/over-representation function.
Added UniProt functions:
rba_uniprot_epitope(), rba_uniprot_epitope_search(), rba_uniprot_coordinates_location_genome(), rba_uniprot_proteomics_hpp(), rba_uniprot_proteomics_hpp_search(), rba_uniprot_proteomics_non_ptm(), rba_uniprot_proteomics_non_ptm_search(), rba_uniprot_proteomics_ptm(), rba_uniprot_proteomics_ptm_search(), rba_uniprot_proteomics_species(), rba_uniprot_rna_edit(), rba_uniprot_rna_edit_search()
Deprecated UniProt functions:
rba_uniprot_proteomics(), rba_uniprot_proteomics_search(), rba_uniprot_ptm(), rba_uniprot_ptm_search()
New arguments added to existing UniProt functions:
rba_uniprot_features_search(), rba_uniprot_coordinates_location_protein(), rba_uniprot_features()
All scripts were reformatted to enhance readability.
Minor improvements and fixes.
Move to JASPAR 2024.
Minor improvements and fixes.
Move to STRING v12.
Incorporate the recent changes in PANTHER and miEAA API.
Minor improvements and fixes.
Submitted the paper to Bioinformatics journal (DOI: 10.1093/bioinformatics/btac172).
Added vignette article: "Over-Representation (Enrichment) Analysis""
Updated citations information.
Moved to JASPAR 2022.
Added new API endpoints of UniProt and PANTHER.
Minor internal improvements.