{"entity":{"id":"idea-tr2-data-link-enforcement","kind":"idea","name":"Journals check that data links actually work, and flag papers whose data vanish","aka":[],"tldr":"Papers say 'data available on request' or link to files that no longer exist. Journals should verify data access at publication and periodically afterwards, and mark papers whose data have disappeared.","summary":"Studies of data availability statements find that most 'available on request' data cannot be obtained, and that links decay within a few years. Automated checks at submission (resolvable repository DOI, non-empty dataset, matching description) and annual re-checks, with an expression of concern for papers whose data become unavailable, would make data sharing real rather than nominal.","asOf":"2026-09-08","links":[{"label":"Bottleneck evidence (Preclinical results do not reproduce): Errington et al., Investigating the replicability of preclinical cancer biology (eLife 2021)","url":"https://doi.org/10.7554/eLife.71601"}],"tags":[],"related":["idea-tr2-raw-image-deposit","idea-tr2-failed-trial-ipd-default"],"cancers":[],"sections":[],"technologies":[],"targets":[],"drugs":[],"companies":[],"institutions":[],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":["b-reproducibility","b-data-silos"],"keyPapers":[],"journals":[],"dependsOn":[],"notes":[],"hypothesis":"Enforcement raises the proportion of oncology papers with actually retrievable data from under 30% to over 80% within three years at adopting journals.","rationale":"Compliance follows verification: sequencing data deposit is near-universal because journals check accession numbers.","test":"Implement at two journals; sample papers annually and attempt retrieval; compare with non-adopting journals.","maturity":"early-clinical","actor":"policy","cost":"small","horizonYears":1},"route":"/ideas/idea-tr2-data-link-enforcement/","neighbours":{"idea":[{"id":"idea-tr2-raw-image-deposit","kind":"idea","name":"Deposit the raw blots, gels and microscopy images behind every figure","route":"/ideas/idea-tr2-raw-image-deposit/"},{"id":"idea-tr2-failed-trial-ipd-default","kind":"idea","name":"Patient-level data from failed trials becomes open by default after two years","route":"/ideas/idea-tr2-failed-trial-ipd-default/"},{"id":"idea-tr2-random-audit","kind":"idea","name":"Random audits of published, funded research, like tax audits","route":"/ideas/idea-tr2-random-audit/"}],"bottleneck":[{"id":"b-data-silos","kind":"bottleneck","name":"Data silos","route":"/bottlenecks/b-data-silos/"},{"id":"b-reproducibility","kind":"bottleneck","name":"Preclinical results do not reproduce","route":"/bottlenecks/b-reproducibility/"}]}}