ZipHelixPrivate by default
This policy applies to ZipHelix, a standalone research product. Your saved projects, private questions, sequences, notes, and account information are not a public research database.
We do not sell your personal information or private research, publish your projects, share them for advertising, or authorize unrelated third-party reuse without your separate, explicit opt-in. If optional sharing is introduced, it must identify the data, recipient or audience, and purpose and remain off unless you choose it. Signing in or agreeing to the Terms is not consent to optional sharing.
Providers needed to operate features you request and limited legal or security disclosures are described below. These are not permission to sell or publish your research. Email team@ziphelix.com for questions or a privacy request.
Internal model development
ZipHelix may use research inputs and outputs to develop, evaluate, and train its internal models. Before introducing training on private research, we will explain the data involved and how it will be used, and obtain any required permission.
Data collected under a no-training promise will not be used for training without separate permission. Internal model development does not authorize sharing identifiable information or private research with unrelated parties.
Information you provide
- Account: email address, sign-in identifiers, and information supplied by your chosen sign-in service. Name and institution are optional.
- Research: project names, selected references and genes, questions, context, results, notes, saved sequences and annotations, simulation inputs, and observations you enter.
- Support: your reply email, optional name, topic, and message when you contact us.
- Request records: which feature you requested and when, the applicable request version, and any separately recorded choices or acknowledgments, where applicable.
Do not submit patient-identifiable data, passwords, access keys, or sensitive personal information the service does not request. Check that you have the rights, permissions, and any necessary consent to process research you provide. Genomic, expression, or health data can remain identifiable even after names are removed; removing a direct identifier alone is not a guarantee of anonymity. Private account storage is not a promise that the service meets a particular clinical, health-record, or institutional compliance standard.
Providing the service
We use information to authenticate you, save and recover your work, run the features you request, respond to support messages, and maintain security and reliability. Hosting and security systems may process request information such as IP addresses, browser details, timestamps, errors, and abuse-prevention records.
Essential providers process the information needed for hosting, account authentication, storage, email, and security. Online account storage and sign-in use Google/Firebase; choosing a sign-in provider also involves that provider. Support messages are handled by the ZipHelix support team through its email providers. Necessary processing is limited to operating and protecting the requested service, not permission for providers to publish or sell your research.
Authorized operational access may be needed for support, security, or fault recovery. Account access controls are not end-to-end encryption, and we do not claim that administrators or infrastructure providers can never access stored data.
ZipHelix AI: information and processing
ZipHelix AI may combine information from ZipHelix’s own knowledge resources, your submitted inputs, and public or licensed research sources. Processing may use ZipHelix-operated technology, third-party AI models and services, or both. Some current features use third-party AI services. The mix may change by feature and over time; this does not change our privacy commitments or permit unrelated use of your private data.
When you choose an AI analysis or research action, you ask us to process the inputs for that request, including its background work and retries. Browsing public references alone does not start an AI research request.
- Gene and outcome research: processing uses the selected gene/reference, proposed functional change, question or biological context, any optional DNA template you provide, and relevant public annotations and publication abstracts.
- Literature searches: species and gene or phenotype search terms are sent to the publication-search service. DNA templates are not included in those search queries.
- Custom reagent analysis: processing uses the submitted reagent details and selected cell, strain, and plasmid annotations. Search providers may receive relevant terms to research those details.
We do not automatically add your account email, profile, account identifier, or project notes to AI requests. Information you type into a question, template, reagent name, or annotation is part of the submitted input, so leave out unrelated personal information.
Service providers receive only information needed for the requested feature and may retain limited request information under their service and security policies. Disabling response storage is not a guarantee of zero provider retention. Requesting an analysis does not authorize public sharing, advertising, sale, or model-training licensing.
Reference lookups and other calculations
Reference searches use public scientific resources such as NCBI. Organism, assembly, accession, gene, or reference-region requests may be sent to the relevant source to return the records you ask for. Publication searches use Europe PMC. Disease and trait searches may send your search terms or public condition identifiers to Monarch to retrieve published associations. Human genetic evidence checks send public gene symbols and condition identifiers to Open Targets. These lookups do not include your account profile, private project notes or edit templates. External links take you to services with their own privacy policies.
For cell-specific DNA comparisons, ZipHelix runs sequence models on its servers. The genome position and submitted reference and alternative alleles are sent to Ensembl for variant annotation; public reference-region requests go to NCBI. Optional variant-ID lookups also use Ensembl. These requests do not include your account profile or project notes.
Regulatory DNA analyses send the submitted sequence, selected changes, and analysis settings to ZipHelix for server-side model calculations.
Drug-sensitivity predictions: uploaded RNA-expression tables are parsed in your browser. The prediction request sends the parsed numeric expression values, normalization setting, and selected compound identifiers to ZipHelix’s servers. The original file name and raw file are not sent by this prediction feature. The prediction handler computes results in memory and does not itself save the submitted profile to an account project or send it to a third-party AI service. Infrastructure may still process request and security records as described above. Reports you download can contain the submitted expression values; keep those files secure.
Some sequence comparisons, visualizations, and calculations run in your browser. Server-based repair forecasts and base-editor scoring send the sequence and analysis inputs needed to ZipHelix’s service. While you are signed in, completed gene searches become projects automatically, and subsequent changes to those projects are saved to your account. Projects may contain your selected references, search terms, notes, research inputs and saved analysis state. You can review and delete projects in your account.
When disclosure may be required
We may disclose the minimum information reasonably necessary to comply with a binding legal obligation or valid legal process, or to investigate and respond to a specific fraud, security, or safety incident. We assess the scope of requests and notify affected users when legally permitted and appropriate. These exceptions do not authorize routine sharing for another organization’s research, marketing, or commercial use.
A change of ownership does not cancel these privacy commitments or authorize new uses of previously collected private data. Any materially different optional use or disclosure requires a separate opt-in before it occurs.
Saving, background work, and browser storage
Signed-out projects can remain in your browser. Signing in saves work to your account and may bring that browser’s guest projects into your account. Browser storage also keeps sign-in state and preferences. Older versions may have stored acknowledgments in that browser; these are not treated as new agreement or permission for unrelated uses.
Research requests you start can continue after you close the page. You can cancel an active request where the control is available; cancellation cannot undo processing already performed or guarantee that an in-flight provider request stops. Completed results linked to a project remain while that project exists. Unfinished, failed, cancelled, and unlinked job records expire after 30 days, subject to cleanup timing.
Retention and deletion
Account projects and optional profile fields remain until you change or delete them. Projects moved to My Data can be restored for 30 days. At the end of that window they become unrecoverable and automatic cleanup removes their contents and linked research results from active storage. Physical cleanup may take additional time.
We retain a minimal project identifier and deletion status to prevent an old offline save from recreating an expired project, without retaining its research contents in that marker. Browser copies are removed when that browser next opens ZipHelix and syncs, or when you clear its site data; we cannot erase an offline device remotely.
Confirmed account deletion removes the authentication account, saved profile, account projects, job records, and deletion markers. It clears that account’s work in the browser completing deletion. Other devices may retain copies until they reopen ZipHelix and sync or their site data is cleared. A short-lived deletion record supports reliable cleanup and retries.
Limited technical records, support correspondence, or protected backups may remain for security, legal requirements, or the providers’ retention periods. They are not a recovery service for deleted projects. Contact support to request deletion of correspondence or other personal information not managed through My Data.
Your controls and rights
Use My Data to review your account information, restore recently deleted projects, and request account deletion. Edit optional name and institution fields through your account. You choose whether to start AI research, can cancel eligible active work, and can delete saved inputs and results. Cancellation cannot reverse processing already completed or legally required retention.
Depending on your location, you may have rights to access, correct, receive a copy of, or delete personal information, or object to or restrict certain processing. Email team@ziphelix.com for questions or a privacy request. We may verify ownership before acting. We do not sell personal information or share it for targeted advertising, so there is no sale or advertising-sharing program to opt out of.
Security, location, and updates
We use access controls and safeguards designed to protect private data, but no online system can guarantee complete security. You are responsible for securing your devices, sign-in methods, and downloaded reports and for deciding whether this service is appropriate for your confidentiality obligations. Report suspected unauthorized access through support. These responsibilities do not reduce our legal data-protection obligations or the commitments in this policy. Providers may process information outside your country. ZipHelix is intended for research users and is not directed to children under 13.
We will date policy changes and provide additional notice when required. We will not treat silence, continued browsing, or a general Terms agreement as opt-in to a new optional use or sharing of your private data.