SIRIUS: vanhemmat versiot ja versiohistoria

- feature: Searchable projects & advanced filtering. Projects now carry a full text search index over their features, which is the basis for the new query and filter capabilities.
- Smart query Bar: Located above the feature list with auto-complete suggestions for fields, operators, and known values (e.g., adduct spellings, lipid class abbreviations, compound class ontologies, and tags). Press Enter to apply. (GUI)
- Unified filtering: Manual panel filters and typed search queries merge into a single filter. The standard filter dialog remains accessible via the funnel icon. (GUI)
- Bulk cleanup: Delete all non-matching features from a project in a single step in the GUI or directly by query in a single API request--eliminating the need to fetch and send back IDs.
- Index management: The search index builds automatically during data import and can be rebuilt anytime from Project Settings if out of sync. (GUI)
- Lucene query syntax: Paged listing endpoints accept searchQuery strings supporting field-specific terms, phrases, ranges, regex, and boolean operators. (API)
- Query metadata endpoint: The searchable-fields endpoint exposes queryable fields, value types, word-search support, and sortability. (API)
- Filter quantification tables directly using search queries. (API)
- Legacy support: Existing CLI feature filter flags (--mzmin, --rtmax, --quality, etc.) remain unchanged and run independently of the search index. (CLI)
- Projects created with older SIRIUS versions automatically migrate and index upon first open.
- Single-click "Restore Defaults" button for feature filters. New single-adduct filter preset.
- feature: Build custom structure databases using reaction rules. Choose from built-in reaction library or create own reactions. Support for multi-product reactions. Inclusion of intermediate products in the final database. (GUI)
- feature: Inspect and browse the contents of custom-built structure databases directly within the GUI.
- feature: Tag objects within a project to group data and compute fold changes. Tagging is used for blank subtraction and to mark PFAS candidates. Custom tag management is currently API-only (these endpoints are still experimental and may change)
- feature: PFAS and special element detection: Integrated specialized PFAS candidate detection automatically detects strong polyfluorination signatures. Elements are now predicted for both de novo formula generation and exclusions in database/bottom-up searches. Detected elements are stored in the project.
- feature: Added dedicated instrument profile for electron-activated dissociation (EAD) QTOF fragmentation tree computation.
- feature: Assign custom sample and run names during LC-MS data import to easily trace results back to input files.
- improvement: Improved GUI responsiveness. Background computations now yield processing priority to the user interface, preventing the GUI from freezing during heavy multi-core processing. (GUI)
- improvement: Upgrading project files from older SIRIUS versions is now faster using all available CPU cores (but may still take a while on large projects). A real-time progress bar replaces previous static screens (appearing frozen). (GUI/API)
- improvement: Online PubChem name lookups are now cached and bound by a timeout, preventing external network outages or slow API responses from stalling searches. (GUI)
- improvement: Added smooth loading animations for embedded web views, enforced minimum web view sizes, introduced borderless icon buttons, and faster rendering for the feature list and its annotations. (GUI)
- improvement: Enhanced filters: Replaced feature quality checkboxes with an intuitive range slider. Results can be sorted by quality. Added a three-state selector PFAS filter. Converted the lipid filter into a clearer three-state selector. (GUI)
- improvement: Significantly improved table loading speeds for quantification tables (API). Made run columns optional. Quantification table endpoints are now stable.
-