RESEARCH RESOURCES
Research Data & Privacy
How AstroNexus stages inputs, publishes results and schedules completed-input cleanup.
Upload & private staging
Your browser uploads input files directly to private Google Drive staging. Inputs are used for your requested analysis. The control plane stores private account, submission, job and file metadata. An authorized worker retrieves inputs into quarantine and performs checksum, container and sequence-format security checks before analysis. Malware scanning depends on the configured worker scanner policy; uploads are not reported as scanned before the worker checks them.
Submission and upload do not require the local compute machine to be online. Once uploaded and queued, progress can be followed from Account or Status without keeping the upload tab open.
Result publication — read before uploading
The current service publishes result reports and curated ZIPs at public links. Anyone with a result URL can view or download it without signing in. Account history and submission details are access-controlled; this does not make the published result links private.
Use non-identifying project/sample labels, since labels may appear in reports. Consider whether public result publication is suitable for your research before uploading sensitive or restricted data. The Demo Center separately lists only explicitly verified public-reference examples.
Raw inputs & delivered results
Source FASTQ reads and source Comparative FASTA assemblies are excluded from public result packages and publication trees by the delivery checks. Generated outputs, including a WGS assembly FASTA, may be included. Public results therefore still contain research information derived from your inputs.
Google authorization credentials, upload-session URLs and worker/operator credentials are kept out of public reports and normal customer-facing logs. These controls are not a certification or a guarantee against every security incident.
Staged-input retention
Original uploaded sequence files are normally retained for up to 7 days after analysis for recovery purposes, then scheduled for deletion. An operator may extend completed-input retention to a maximum of 14 days. Cleanup is retried and can be delayed by service outages. Active and queued inputs are protected.
Failed/cancelled inputs have a 72-hour retention policy, but stay protected while recovery needs them. Explicitly registered test inputs are cleaned after their assertions pass. Private originals are organized by job; archive links are not public. The seven-day rule does not delete published results or guarantee deletion of every copy. This describes implemented research-data handling, not a legal compliance policy.
Research use only. Not for clinical or diagnostic use.