Run to BioSample, via the ENA portal

Route id: run->biosample:ena_filereport Provider: EBI ENA portal Direction: SRR/ERR/DRR → SAMN/SAMEA/SAMD

curl -s 'https://www.ebi.ac.uk/ena/portal/api/filereport?accession=SRR25056225&result=read_run&fields=sample_accession&format=tsv'

The BioSample is the identity that crosses archives. GEO knows a sample as a GSM, SRA knows it as an SRS, and the BioSample is the one both record — which is what makes it the join key when GEO omits the SRA relation entirely (the recovery that attributes 36 of 36 runs for GSE135325).

Measured

Corpus: sample:3000@reprocessed-srr. Surveyed: 2026-09-15.

Resolved	2,977
Empty	23
Failed	0
Unique BioSamples	2,880

Identical resolution to run → experiment, and for the same reason: both come from one filereport row, so a run either has a row or it does not. The 23 empties are exactly the same runs, and they are absent from both archives now.

2,880 BioSamples for 2,977 runs — slightly more sharing than at the experiment level (2,888), which is what you would expect: one biological sample can carry several libraries.

One request, several answers

run_accession, experiment_accession, sample_accession, secondary_sample_accession, study_accession and secondary_study_accession all come from the same row. Declaring them as separate routes is about being able to rank and measure each edge separately; it is not about issuing separate requests. fetch relations fetches the row once and fills every field from it.