Privacy Policy

1. Scope and principles

This policy applies to SeqBench (https://seqbench.com), including the website, its programmatic interfaces, account services and related support. It should be read together with the Terms of Service.

The core tools can be used without creating an account or providing credentials. The service is built for research, education and general molecular-biology workflows; it is not intended for Protected Health Information or other patient-identifiable clinical data and must not be used as part of a clinical diagnostic process.

2. Accounts and sign-in

If you choose to sign in, an external identity provider authenticates you. SeqBench never sees your password. The provider returns your account identifier, email address, name and profile picture URL so the service can recognise you on a later visit; no broader access to your account is requested. Signing in also tells that provider that you signed in to SeqBench.

You may sign out from the account menu at any time, which removes the session cookies from your browser. To delete the account record itself, contact us using the details below.

3. Sequences, files and project data

Sequence data is not stored by default

Sequences and files submitted through the website or its programmatic interfaces are processed to compute a result and are not written to a database, logged or retained afterward, with three explicit exceptions:session_* tools keep named scratch values for 24 hours;save_permalink permanently stores an anonymous input/result pair under a short unguessable code so anyone with the link can reopen it; and the four modelling services — deep plasmid annotation, promoter strength, RBS strength and pegRNA scoring — keep each result, together with the sequence it was computed for, in server memory for up to 24 hours, so an identical repeat request is answered without recomputing. That copy is never written to disk and is discarded when the service restarts.

Tools that retrieve public records send only the public identifier needed to fetch that record, and keep the record they fetched in memory for at most an hour so a repeated lookup does not query the source again. Server-side multiple-alignment modes keep sequences only for the duration of the run and remove them when it finishes, fails or is cancelled. They are not sent to an external alignment service, and a browser-local alignment remains available without uploading anything.

What stays in the browser

Most workbench activity does not reach a server: editing, restriction digests, the virtual gel, translation, primer design, the cloning simulators, plate-scale clone verdicts, the folder library, primer inventory and .dna/GenBank import and export all run in the browser. A folder you adopt is indexed where it sits; its files are read locally and are not uploaded.

Fifteen workbench capabilities do call a server, and each identifies that fact in its own panel: deep annotation, plasmid identification, Golden Gate junction fidelity, CRISPR off-target screening, the CRISPR HDR donor, RBS design, construct verification, Sanger readback verification, fetch-by-accession, pegRNA ranking, plate-scale editing quantification, protein domain submission, the poll that collects that job's result, identifying a selected region against GenBank, which sends that region to NCBI, and the cryptic-promoter scan. Domain annotation is counted as two calls: the residues leave on submission, and the job identifier leaves again on each poll until the job finishes.

Browser-local file readers—including Sanger chromatograms, SnapGene .dna files, GenBank records and format conversion—decode files in the tab. A workflow that explicitly sends a selected trace or derived sequence to the server says so at its input. Everything else is computed on the server under the transient-processing terms above.

Local project history and optional sync

Tool results and their inputs are written to this browser's local database so My Projects can reopen prior work. By default, that history is not uploaded, cannot be read by SeqBench, does not follow you to another browser or device, and can be removed from the Projects page or by clearing this site's data. The maps you compare sequencing results against (or check an oligo order against) are kept the same way—in this browser only, the last twelve—so the next comparison can offer them back, and so a new result can be matched to the map it belongs to. That matching happens in this browser; only the map it picks is sent, for the comparison itself, as when you chose it by hand. Each can be removed with its ×. To keep none of this, switch off “Save in this browser” on My Projects: from then on no result, workbench session, file or map is written to this browser (what is already there stays until you delete it).

Cloud sync changes this only when you ask for it. If you sign in, you may choose to copy the records on My Projects to your account so another computer can open them. Sync is off until you select it. The newest 300 records are copied — their inputs, names and findings, and the files saved with them, such as sequencing traces and reads. Files are kept in private storage that only your account can read, and leave the account when their record does. Adopted folders and their files are not copied. You can stop sync at any time and separately choose whether to delete the account copy, files included. Neither action removes local data from your devices.

4. What is logged

To monitor reliability and understand which tools are useful, each run records usage metadata: the tool called, whether it succeeded, run time, input/output size and a client identifier such as an MCP client or user-agent name. A run may also record the failure category, whether it produced results, how the input arrived and—when a file could not be read—the detected format and the file extension. This metadata never includes a sequence, file contents, file name or other submitted data.

Runs made from the website carry a pseudonymous identifier so SeqBench can distinguish one browser making several runs from several browsers making one. It is not a name, email address or account ID, is not shared, and is never used to decide access. Runs made through the REST API or MCP server carry no visitor identifier. Hosting and network providers may retain standard connection-level logs, such as an IP address and request timestamp, for a limited period for security and abuse prevention.

5. Files deliberately sent for parser support

When a file cannot be opened, the error may offer a way to send a sample so the reader can be improved. Nothing is uploaded when the file is refused or when the form is opened. Upload occurs only after you press the button that explicitly says the file will be sent.

If you proceed, SeqBench receives the first 64 kB of the file, its extension, total size, the page that refused it, and anything you choose to enter in the note and email fields. The file name is not sent. These samples are kept privately, used only to fix and test file readers, and are not published or shared. Contact us to request deletion of a submitted sample.

6. Cookies and analytics

SeqBench uses privacy-focused, cookieless analytics to understand aggregate traffic, such as page visits and overall visitor counts. It does not track people across other websites, and analytics traffic never includes a sequence, file or other data submitted to a tool.SeqBench does not use advertising trackers and does not sell or share data with advertisers.

To count visits rather than page loads, your browser stores a random identifier. It is not a cookie, is sent only to SeqBench, does not follow you to another browser or machine, and clearing this site's data ends it. If you sign in, the identifier is instead derived from your account ID by a one-way hash, allowing the same person to be counted across their own devices without sending the account ID or email address to analytics.

Signing in records that the anonymous browser identifier and signed-in identifier belong to the same person, so activity before sign-in is not counted as a separate visitor. That record contains the identifier pair and the date it was first made—no page, tool or submitted data—and is used only for service measurement. Clearing site data before signing in leaves nothing to link.

Cookies are used only to keep a signed-in user authenticated. These session cookies are necessary for account access, cannot be read by browser scripts, are not used for advertising and are removed when you sign out.

7. Donations

Donating is optional. If you donate, the payment is completed on a checkout page hosted by our payment processor, which collects your payment details and email address directly under its own privacy policy; SeqBench never receives your card details. The processor also receives the amount and, if you enter them on the donation form, your name and email address. If you are signed in, your account identifier is attached so early access can be given to your account.

Donation records are used only to process and account for donations, to answer questions about them and to invite donors to features in early access. They are not linked to the usage records described above.

8. Service providers and external sources

SeqBench uses selected providers for cloud hosting, managed data storage, authentication, payment processing and aggregate analytics. They process information only to provide those functions to SeqBench. Signing in involves the identity provider under its own privacy policy. Tools that look up public records — sequence accessions, genes, variants — contact the relevant public database with only the identifier or variant description supplied. A similarity search — the BLAST tool, or identifying a Sanger read or contig — is the exception: when you press its button, the bases being searched, and no file or sample name, are sent to a public sequence database's search service, which handles them under its own terms. Submitted sequence data is not sold or shared with advertisers. Contact us for the current list of processors.

Information may be processed in countries other than the one in which you live. Where applicable law requires it, recognised safeguards are used for international transfers.

9. Your choices and rights

If you have an account or otherwise believe SeqBench holds personal information about you, you may ask us to review, correct, export or delete it, subject to applicable law. You may also have rights to object to or restrict certain processing and to complain to a local data-protection authority.

We may need to verify your identity before completing a request. Some information may be retained where required by law or reasonably needed to prevent abuse, resolve disputes or establish, exercise or defend legal claims.

10. Changes and contact

This policy may be updated as the service or applicable requirements change. The date at the top identifies the latest revision. If a change materially affects how personal information is handled, we will take reasonable steps to provide additional notice.

Privacy questions and requests can be sent to seqbench.bio@gmail.com. Please do not include sequence data or confidential project details in the initial message.