Text mining and analysis of "data available upon request" statements in scientific articles. (Data)

This Zenodo record contains the data and data tables used to analyse the use of “data available upon request” style statements in the genomics, genetics, and bioinformatics literature (2010–2025).

This dataset accompanies the corresponding code repository (GitHub) and the automatically archived code release on Zenodo (Code - Text mining and analysis of "data available upon request" statements in scientific articles.).

The analyses quantify how often the "upon request" wording is used, how it changes over time, and how it relates to other signals of open science support at the article and journal level.

Overview

Using full-text XML articles from the PubMed Central Open Access Subset, the pipeline detects “upon request” sentences, classifies them (vague versus linked to explicit access mechanisms or legitimate restrictions), and extracts several open science indicators (for example public repository deposition, dataset identifiers, code availability signals, protocol sharing, and source data mentions). The tables in `2.data/` are the inputs and derived metadata used to run the analyses and reproduce the manuscript figures and summary statistics.

Folder contents

• 2.data/
  Data tables used by the pipeline.

Main files:
- `meta_all_papers.tsv`  
  Article-level metadata for the full journal set and time window.
- `meta_openaccess.tsv`  
  Subset restricted to PMC Open Access Subset articles.
- `meta_openaccess_enriched.tsv`, `meta_openaccess_enriched2.csv`  
  Enriched metadata tables used during downstream processing (see code for details).
- `meta_under_request.tsv`  
  Articles with detected “upon request” wording.
- `meta_under_request_tagged.tsv`  
  Tagged classification of “upon request” statements (article-level summary fields).
- `meta_under_request_tagged_open_science.tsv`  
  Article-level table including open science indicator signals and OSSI-related fields.
- `meta_under_request_tagged.xlsx`  
  Convenience spreadsheet version of the tagged table.
- `metadata_ids_all_papers.txt`  
  ID list used to track the corpus.
- `phrases_that_have_marched.txt`  
  Matched phrases / patterns used for detection and QC.
- `policy_timeline.tsv`  
  Curated timeline of relevant policy and reporting changes.

Subfolder:
- `pmc_ftp/`  
  Reference inventories from PMC used to define the Open Access Subset:
  - `oa_file_list.csv.zip` (official PMC OA file list snapshot used for filtering)
  - `PMC-ids.csv.zip` (supporting inventory)
  - `README.txt`

License

Software. The analysis code is released under the license provided in the associated GitHub repository (`LICENSE`).

Tables and derived files produced in this project. Reuse of the tables and derived datasets created by the author of this record follows the license selected for this Zenodo record (see Zenodo metadata).

The license selected for this Zenodo record applies only to the files produced in this project (tables) and does not modify or override the license of any XML (included in another Zenodo record). 

Citation

If you use these XML files in academic work, please cite the associated manuscript and the Zenodo records below (code, XML corpus, and derived tables).

Manuscript: 
Ballester, B. (2026). *From ‘data available upon request’ to accountable data access in genomics*. DOI: to be added.

Code (Zenodo):
Ballester, B. (2026). *Code: Text mining and analysis of “data available upon request” statements in scientific articles* (v1). Zenodo. https://doi.org/10.5281/zenodo.18339878

Full-text XML corpus (this record):  
Ballester, B. (2026). *XML: PubMed Central Open Access Subset JATS XML files used for “data available upon request” analyses* (2010–2025). Zenodo. DOI: https://doi.org/10.5281/zenodo.18377386 

Derived tables (2.data record):  
Ballester, B. (2026). *Data: Text mining and analysis of “data available upon request” statements in scientific articles*. Zenodo. DOI: https://doi.org/10.5281/zenodo.18375259 

Contact

Benoît Ballester (Aix Marseille Univ, INSERM, TAGC, UMR 1090, Marseille, France)