Evidence for this page

Back to the page

In one local run of the edge variant on 20 October 2026, a SARS-CoV-2 genome file (FASTA, NCBI), a synthetic alignment (BAM) and a synthetic variant file (VCF) were stored and retrieved through DRS. The SHA-256 checksums were equal. Only the genome file was encrypted and decrypted with Crypt4GH. Its SHA-256 checksum was equal. The two synthetic files were not encrypted or decrypted with Crypt4GH.

Tested

Date: 20 October 2026

The alignment and variant file were synthetic, not a real dataset. Regions through htsget are not possible. This is not a statement about Beacon for pathogen data.

Evidence if you ask

Ferrum stores Nanopore raw files (POD5, FAST5, BLOW5) and quality metrics supplied by the caller.

Tested

Date: 14 October 2026

Ferrum does not read these files itself and computes no metrics from them.

Source code Evidence if you ask

A partner can ask for access to a dataset. For a file bound to a dataset, Ferrum allows access when the sign-in carries a release for that dataset or when the caller is an administrator. Ferrum does not decide the request.

Present in the code

Date: 7 November 2026

Present in the code, no end-to-end test. The request exists only when an external decision service is configured. The data holder's decision is not in this program. Unbound files and an administrator account are outside this check.

Source code Source code Source code Source code

Collection date, place and host are not fields you can query.

Tested

Date: 24 October 2026

An organism can be stored and queried. Collection time, place, coordinates, country, host species and an attached metadata object can sit in other records or in a bundle, but they were not part of the sample query that was checked. The metadata interface stays off until it is switched on, this registration did not use the bundle, and the proposed table for pathogen samples is not among the database changes.

Source code Evidence if you ask

A separate Beacon for pathogens is decided, not built. The human-focused Beacon rejects the SARS-CoV-2 reference name NC_045512.2 with HTTP 400.

Tested

Date: 24 October 2026

This name was checked, not every pathogen reference. An organism name can still be queried on the human Beacon. The separate Beacon is not built. This is not a GA4GH certificate.

Evidence if you ask

htsget in Ferrum currently returns whole files. Regions (reference name, start, end) are rejected for every kind of data (HTTP 400).

Present in the code

Date: 20 October 2026

A trial showed that the library htsget-search 0.14.0 produces region tickets for BAM and VCF. It is not built in yet. The edge size against the 52,428,800-byte gate is not in the evidence file.

Source code Source code Evidence if you ask

No client for ENA or Africa CDC platforms exists.

Present in the code

Date: 24 October 2026

A local archive writes a file and does not upload it. No national or continental network was contacted.

Source code Source code Evidence if you ask

According to Australian developers (CSIRO), the GA4GH Beacon protocol is focused on human data and has to be adapted for data on pathogens. Adaptations such as PathSBeacon exist. The Public Health Alliance for Genomic Epidemiology (PHA4GE) ran a workshop in 2023 on that. Only Beacon is discussed here. The file-level standards DRS, htsget and Crypt4GH were not assessed here.

Source

Date: 21 October 2026

This entry assesses the Beacon part. The file-level standards DRS, htsget and Crypt4GH were not assessed here. No separate pathogen Beacon is built. The file run of 20 October 2026 is a different entry.

Source Source Source