EBI — ArrayExpress SDRF files point at a decommissioned mirror

Status: draft, not filed Archive: EBI ArrayExpress / BioStudies Re-checked: 2026-10-15 — both studies reproduce as above (E-MTAB-8060: 2 files, 0 fastq; E-MTAB-9221: 43 files, 40 fastq), and the pathology test passes live. Evidence: pathologies/ae-sdrf-points-at-decommissioned-mirror.md Where to file: biostudies@ebi.ac.uk

for AE in E-MTAB-8060 E-MTAB-9221; do
  printf '%s: ' "$AE"
  curl -sL "https://www.ebi.ac.uk/biostudies/api/v1/studies/$AE/files?limit=500" \
    | python3 -c 'import json,sys; i=json.load(sys.stdin)["items"]; print(len(i),"files,",sum(1 for f in i if ".fastq" in f.get("path","")),"fastq")'
done

---

Draft report

Subject: Comment[FASTQ_URI] in SDRF files points under /pub/databases/microarray/, which is no longer maintained

Many ArrayExpress SDRF files carry Comment[FASTQ_URI] values under

ftp://ftp.ebi.ac.uk/pub/databases/{microarray,arrayexpress}/data/experiment/...

the pre-BioStudies mirror. The same spelling means two different things and the SDRF gives no way to tell them apart:

• the file still exists under BioStudies and the URI is a stale spelling of a live study file; or
• the file does not exist anywhere, and the study's actual data is elsewhere.

E-MTAB-8060 and E-MTAB-9221 are indistinguishable at the SDRF level. Both point every Comment[FASTQ_URI] at the mirror; both also declare a Comment[BAM_URI]. Yet:

	Files the study registers	of which fastq
E-MTAB-9221	43	40
E-MTAB-8060	2 (the IDF and the SDRF)	0

E-MTAB-8060's reads exist only on the unmaintained mirror; its real submission is the ENA BAM. A client taking the SDRF at face value discards that BAM for all 15 runs, with nothing anywhere recording that it did so.

What would resolve it

Rewriting the URIs to their BioStudies paths where the file is still registered would fix the first category outright. For the second, a study whose SDRF references files it does not register is a data-integrity signal worth surfacing at submission time.

We currently work around it by checking each mirror URI against the study's registered file list and dropping those that do not appear, which is only possible because the file listing exists. It is not something every consumer of an SDRF will think to do.