An ArrayExpress IDF does not declare the ENA study it has

Archive: EBI ArrayExpress / BioStudies Affected route: ae_experiment->study:idf_secondary Status: measured; partly worked around by a fallback; reproduced 2026-10-23 Measured: 2026-09-15, every ArrayExpress study BioStudies records an ENA link for

The number

551 of 20,693 studies (2.7%). That is the number of studies, among every ArrayExpress study whose BioStudies record carries an ENA link, whose IDF declares no Comment[SecondaryAccession]. The IDF is the field that names the ENA study. The link exists, and the file that is supposed to state it does not.

Reproducer

curl -s https://www.ebi.ac.uk/biostudies/files/E-MTAB-6505/E-MTAB-6505.idf.txt | grep -ci 'SecondaryAccession'
# 0
# while its SDRF's Comment[BioSD_SAMPLE] SAMEA5053920 resolves, through ENA, to PRJEB29431 / ERP111731

Checked on 2026-10-23.

What it is not

• Not a missing study. The ENA study exists, and BioStudies links to it. Only the IDF is silent.
• Not the 5 IDFs that answer 404. Those studies register no files at all, and are counted as failures, not empties.

Workaround

Fall back through the SDRF's BioSamples:

1. ae_experiment->biosample:sdrf;
2. then biosample->run:ena_filereport.

fetch_series.relations walks this composition for any ArrayExpress entry point. Measured over exactly the 551 (corpus arrayexpress-no-secondary), the fallback recovers 4.2%. The route page has the full table.

Affected accessions

The 551 are the committed corpus data/corpora/arrayexpress-no-secondary/, and the EMPTY verdicts of ae_experiment->study:idf_secondary in data/surveys/arrayexpress-ena/.

Regression test

tests/pathologies/test_ae_idf_missing_secondary_accession.py asserts that E-MTAB-6505's IDF still declares no secondary accession.