Data Contribution and Use Agreement

This agreement covers the iSynth data platform and iSynth Notebook, its electronic laboratory notebook (ELN) service: planning, projects and groups, experimental operations and measurements, imports and exports, collaboration, dataset contribution, search, construction and use. Plans and private notebook saving do not constitute public contribution. A publication request requires separate confirmation of the relevant content, provenance and authorization.

Existing rights remain with contributors or original rights holders; uploading does not transfer copyright. Authorship must reflect actual research or creative contributions. Uploaders, curators and iSynth do not automatically become authors of the original data. iSynth may hold applicable rights in its independently created, protectable curation and arrangement, without exclusive rights in scientific facts or third-party data.

All submissions, including drafts, must truthfully describe their sources and nature, distinguishing experimental observations, literature compilations, computed results and artificially constructed test data. Retain verifiable references, original records or computational methods, and disclose known gaps, limitations and processing. Do not fabricate experiments, alter results deceptively, invent sources or impersonate authors. Failed experiments and negative results are eligible; integrity does not require a successful reaction. Artificial workflow test fixtures must be labeled and must not enter the public reaction collection as experimental evidence or receive production DOIs. Automated checks and publication review do not guarantee scientific conclusions. Correct discovered errors promptly. Distinguish experimental plans, unmeasured outcomes and observed non-detection or negative results; a missing yield is not a zero yield.

You confirm that you may submit the data and grant iSynth, to the extent you may authorize it, a worldwide, non-exclusive, royalty-free license to store, back up, validate, clean, standardize, correct, calculate derived attributes, index, deduplicate, compile and create derived datasets, and to provide access and redistribution under applicable licenses. Service providers bound by confidentiality and purpose limitations may perform necessary technical processing. The platform must retain sources and change records and distinguish reported observations from calculated or curated results. This authorization is limited by your selected access scope; it does not authorize automatic publication of private data or their use for platform model training or self-evolution.

Drafts and restricted data remain within their access permissions. Approved public data may enter the iSynth reaction collection for search, record-level display, filtering, combination and dataset construction through the website, downloads, API and MCP. Individual records and constructed datasets retain links to their source datasets and versions. These permitted operations do not require a separate request to the contributor each time. Unlicensed full papers, figures, personal information and other third-party material must not be published with the data.

Saving, importing or creating a dataset from notebook entries does not automatically publish it. Project owners may grant verified signed-in accounts view, comment or edit access to a project or group. Supervisors have no automatic access; collaborators cannot manage sharing or publish data. Dataset account sharing provides read-only file and metadata access to approved published versions, excluding new drafts, offline content and trash; authorized readers may still participate in discussions under platform rules. Owners can send invitations. An invitation grants no access until the recipient completes normal academic admission, verifies the invited email and accepts while signed in. Owners can inspect and revoke account grants and pending invitations. Account grants differ from password links: links or codes can be forwarded, so the platform cannot identify every holder. Owners and administrators may disable, reset or restore code access within their management permissions; disabling or resetting invalidates earlier credentials. Administrators may inspect sharing information and revoke grants or invitations; they do not create account grants on an owner’s behalf. Administrative access and actions are retained in internal security audits; owners can inspect permission-change records needed to manage sharing. Revocation cannot recall downloaded copies. Sharing changes neither ownership, applicable licenses nor publication status.

AI assistance is invoked by the user and uses the personal configuration or site-default model service shown in account settings. Experiment planning sends the objectives, plan description and count you enter. Description extraction and bulk-edit interpretation send the entered description. Analysis and writing send the scientific content of selected experiments and the question after confirmation, including structures, conditions, measurements and objectives; they do not automatically send the whole project or attachments. A third party may operate the model service; check its retention, processing and transfer terms before use. Configuring or invoking a model is not permission to publish data or train on them. Researchers must review generated conditions, suggestions and prose; they must not be represented as performed experiments or measured results. The platform currently does not automatically train models on private experiments or run a self-evolving system.

Private saving requires no publication review. Only a new public DOI requires preliminary administrator review, followed by manual export and external review. An administrator confirms external approval in iSynth before completing that publication and marking the DOI registered. Preliminary review, suffix reservation and export do not constitute registration. Metadata corrections, restricted releases, and republication or restoration with identical file content reusing a confirmed DOI need platform review only. Registration-agency metadata updates may be submitted manually without another publication confirmation. An administrator configures the allocated numeric prefix; new version DOIs use independent opaque suffixes of the form 10.numeric/isynth.{12 lowercase Crockford Base32 characters}; already reserved or registered DOI names remain unchanged. Version numbers and relationships belong in metadata, and all releases stay grouped on one dataset page. Do not derive other DOIs by editing a suffix. Only changes to the collection of file contents, such as additions, removals or replacement by different content, may create a new DOI file release. Description, author or other metadata edits, renaming, reordering and identical uploads do not increment the DOI. Review does not automatically publish restricted data or assign a DOI. Cite a confirmed dataset DOI using its full https://doi.org/ link; otherwise use the stable dataset-version URL.

New original contributions you own or are expressly authorized to relicense default to CC BY 4.0, permitting copying, curation, adaptation and redistribution, including commercial use. Public sharing requires appropriate attribution, license information and a link, and change notices. CC BY 4.0 does not require an entire newly constructed dataset to use the same license, but recipients must remain able to exercise the licensed rights in the original material. Changing the default neither changes existing data-version licenses nor expands previous authorizations. Verify permission to submit and redistribute third-party data and retain its source license. CC BY-NC, CC BY-SA and other sources retain their conditions; mixed sources do not automatically become uniformly CC BY. Restricted or unclear permissions require appropriate authorization before public release.

When research uses a complete dataset or a subset, cite the authors, title, version and DOI of the datasets actually used, or a stable URL if no DOI exists. Filtering, cleaning, merging, annotating or constructing a new dataset does not automatically remove attribution or license obligations for the original material. Shared results should retain source datasets, authors, versions, licenses and change notices. Numerous sources may be listed reasonably in an appendix or accompanying citation manifest; complete citations need not be repeated in every row. Identify your curation or added data without presenting others’ data as independently generated. Acknowledge iSynth search or construction services where appropriate, but citing the platform or new dataset does not replace attribution required by source licenses. Scholarly citation and license obligations are distinct: citation remains recommended for CC0 and facts lawfully used without permission, without imposing additional legal restrictions on them.

The website, API, MCP and ReactionSeek use the same acceptance rules. Automation may transmit only the user’s explicit authorization. The platform records the account, time, agreement version and associated data version or build. Report errors or rights claims to isynth@ichemdata.com; the platform may correct, restrict or remove content. Withdrawal does not revoke previously granted irrevocable licenses or recall lawfully acquired external copies. Deleted unpublished material is retained only for necessary backup, security, audit or legal retention. Public-domain status, statutory exceptions and mandatory law remain unaffected. The platform cannot guarantee or comprehensively monitor external use after download. Licenses apply only to rights that the licensor may lawfully grant and do not create exclusive rights in scientific facts.