Release Date:
Summary
Stack 609 lays the groundwork for smarter, meaning-based search in data portals and delivers the first production pieces of flexible, form-based access requirements. Alongside those, it brings substantial reliability and clarity improvements to the Curator data grid, privacy-protected views in Cohort Builder 2.0, and a set of quality-of-life fixes for electronic Data Use Certificates (eDUCs) and everyday Synapse use.
Key areas of work in this release include portal search (semantic and hybrid search, advanced query syntax and index reliability), data access governance (form template authoring, save-and-resume requests, eDUC preview and flexibility), metadata curation (grid stability, loading feedback and CSV upload accuracy), cohort building (privacy safeguards and cohort hand-off between views), and security hardening.
In total, 38 tickets were completed across the Synapse platform, web client and portals.
No breaking changes for end users. One configuration requirement applies to data managers and ACT members who set up aggregate (count-only) data access; see Deprecated or Breaking Changes below.
Affected Users & Systems
|
Who / What |
Impact |
|---|---|
|
Data portal users and portal teams |
Foundation for search that understands meaning as well as keywords, more powerful search syntax, and more dependable search index builds. |
|
Access & Compliance Team (ACT) |
New tools to build, preview and manage form templates for access requirements; more flexible eDUC templates and fewer eDUC errors. |
|
Data requesters |
Can save partially completed access requests and return later, preview eDUCs before sending, and see clear email validation messages. |
|
Data contributors and curators (Curator) |
A more stable grid, clearer loading messages, and more accurate CSV uploads. |
|
Cohort Builder users and data managers |
Privacy-protected file views, the ability to carry a cohort from one view to another, and easier identification of IDs. |
|
All Synapse users |
Table display, Terms of Use, chat assistant placement and security fixes. |
|
Platform infrastructure |
Deployment tooling improvements with no user-facing impact. |
Deprecated or Breaking Changes
There are no breaking changes for end users in this release.
Quasi-identifier columns now required for aggregate data access settings
As part of the new privacy safeguards in Cohort Builder 2.0, any aggregate (count-only) data access configuration must now list the columns that could identify a participant (known as quasi-identifiers). The platform uses this list to make sure participant information is only ever exposed as counts, and that very small counts are hidden.
Action required: Data managers and ACT members configuring aggregate data access should specify the quasi-identifier columns for the associated participants table when creating or updating a configuration.
Who benefits: Study participants, whose privacy is better protected, and data managers, who get an explicit, auditable privacy setting.
New Features
Smarter portal search: meaning-based “semantic” and hybrid search New
The search engine behind data portals can now understand the semantic meaning of a query, not just match its exact words. Portal teams can choose which fields are searched by meaning, and results from keyword and meaning-based matching are blended into a single ranked list, while still respecting each user's data access permissions. Teams can also save reusable relevance-tuning settings and apply them across multiple portals rather than configuring each one separately. Portal teams will need to provide field names for semantic search in order to have this feature enabled in their respective portals.
Who benefits: Portal users looking for studies or datasets who may not know the exact terms used, and portal teams who want better search relevance with less configuration work.
Advanced search syntax for portal search New
Portal search now supports a richer query syntax, including AND/OR/NOT logic, grouping, ranges, wildcards and approximate (fuzzy) matching in a single search expression. This gives power users and portal builders much finer control over what a search returns.
Who benefits: Experienced researchers and portal developers building precise or advanced search experiences.
Build and preview access requirement form templates New
ACT members can now create form templates for access requirements: they define the questions, arrange them into steps, add guidance, and preview exactly what requesters will see before saving. A new Form Templates tab in the Access Requirement Dashboard makes it easy to find, search and manage these templates. All templates now build on a shared standard foundation, which keeps them consistent with one another.
Who benefits: ACT members, who can tailor access forms to each dataset without engineering help, and requesters, who get clearer, purpose-built forms.
Save and resume data access requests New
Requesters filling out the new form-based access requests can save their progress and come back later to finish. This also allows the new form-based requests to work with the existing eDUC signing process.
Who benefits: Researchers completing longer access requests that require gathering information from colleagues or institutions.
Preview eDUCs before sending, with more flexible templates New
Data access requesters can now review exactly how their electronic Data Use Certificate will look, with their information filled in, before it is sent out for signature. eDUC templates also support a wider range of field types, including pre-filled fields that signers can see but not change.
Who benefits: Requesters, who can catch errors before sending, and ACT members, who have more freedom in designing eDUC templates.
Privacy-protected file view in Cohort Builder 2.0 New
Cohort Builder 2.0 can now show a list of files for a cohort, each with the number of associated participants, while keeping participant-level data protected. Participant details are only ever shown as counts, and any count small enough to risk identifying someone is automatically hidden. Behind the scenes, access checks now run against exactly the data that is being served, making permissions more consistent.
Who benefits: Researchers exploring controlled cohorts, and data managers responsible for participant privacy.
Carry a cohort from one view to another in Cohort Builder 2.0 New
Users can now define a cohort in one view (for example, participants who have both a certain type of file and another) and then browse the matching data in a different view (for example, a list of files) without losing the cohort's meaning. Previously, conditions that made sense for participants could return nothing when applied to files.
Who benefits: Researchers who build a cohort around participants and then want to find and download the related files.
Add RecordSets to File Views New
Users can now include a RecordSet in a File View's scope directly from the Synapse interface, and see which RecordSets are included.
Who benefits: Data managers and portal teams who want to offer complete metadata records for download alongside data files.
Fixes & Improvements
Clearer loading and validation status in the Curator grid
Large datasets can take a few minutes to load and validate. The grid now shows a clear "Preparing your grid" screen with a note that large datasets may take up to five minutes, and better indicates when validation is still in progress. This prevents users from refreshing unnecessarily or assuming their data is missing or wrong.
Who benefits: Data contributors working with large metadata sets.
More reliable saving and syncing in the Curator grid
Edits in the grid are now sent and confirmed one at a time, so changes arrive in the correct order. This fixes cases where edits from one person could be applied out of sequence or missed by other people viewing the same grid. The grid also now recovers if column information briefly gets out of step during syncing, and an incorrect validation error on valid text values was resolved.
Who benefits: Teams collaborating on the same metadata, especially on large or busy grids.
More accurate CSV uploads to the Curator grid
Several issues with CSV uploads have been fixed: values starting with "syn" are no longer mistaken for Synapse IDs when the schema expects plain text; empty optional columns no longer trigger validation errors; and values that look like dates (such as version labels) are no longer converted into numbers.
Who benefits: Data contributors uploading metadata spreadsheets
More dependable portal search indexes
Search indexes now rebuild correctly after the underlying data's columns change, instead of failing or misaligning data. Search also now applies the intended text-matching settings by default, and paging through results consistently returns every page of results.
Who benefits: Portal users and portal teams relying on accurate, complete search results.
All table columns now display correctly
Fixed an issue where some Synapse tables only displayed their first seven columns, even though all data was present.
Who benefits: Anyone viewing wide tables in Synapse.
Smoother eDUC experience
Requesters who enter an invalid email address in an eDUC now see a clear error message instead of being silently sent back to the first screen. Voiding an eDUC that had already been cancelled no longer causes an error, so users can restart the process cleanly.
Who benefits: Data access requesters and ACT members managing eDUCs.
Terms of Use acceptance no longer gets stuck
Fixed an issue where some users were caught in an endless loop when trying to accept updated Terms of Use, which blocked them from continuing into Synapse.
Who benefits: All users accepting updated Terms of Use.
Chat assistant no longer blocks page content
The floating chat assistant has been adjusted so it no longer covers buttons and other elements in portals and Curator.
Who benefits: Portal and Curator users.
Curator Dashboard settings open at the top
Opening Settings from the Curator Dashboard now always starts at the top of the page, rather than inheriting the dashboard's scroll position.
Who benefits: Curators managing task settings.
Easier to match IDs and names in Cohort Builder
Cohort Builder now makes it easier to see both the name and the Synapse ID of items such as parent folders, so users can match conditions built on IDs with the names shown in results.
Who benefits: Cohort Builder users filtering by folder or other Synapse IDs.
Security hardening
The Synapse web client was reviewed and updated to close off potential security weaknesses in an older user-interface library and in a third-party software component. There is no action required from users.
Who benefits: All Synapse users and the organizations that depend on Synapse's security and compliance.
Behind-the-scenes stability and deployment improvements
Fixed a harmless error that could appear in the browser console on page load, resolved an issue affecting developers running Synapse locally, and made the platform's deployment process more resilient so releases complete reliably. These changes have no user-facing impact.
Who benefits: The Synapse engineering team, and indirectly all users through smoother, more reliable releases.