2026.1 Release Notes
Release Highlights
File upload via drag/drop has been improved. You can now drag a file directly into the Data or 3D & Analyze pages and the upload will occur without an upload box. A banner at the top of the window indicates when the upload is finished, and you can also watch the progress in the Jobs tab, although you must toggle on system jobs to see it. In the Data page, the dataset will not be activated automatically, but it is immediately active on the 3D & Analyze page if the file is uploaded there. A toggle on the Preferences tab of the user profile controls this setting: the default is toggled on.
Owners of a custom database and those with delete permissions granted via OCLI may delete the database from its card in the Databases tab of the System page.
User Interface Notes
3D & Analyze Page
The plot legend previously showed a point on the legend for all active datasets, even if there was not a point on the plot from that dataset. The legend now shows the message “No active data in plot” if that dataset does not display any points on the plot.
In the plot settings, there is now a drop-down option for the ‘violin’ and ‘boxplot’ plot types, in which a user can choose the plot’s orientation as auto (default), horizontal, or vertical.
A larger set of results can now be handled by the 3D & Analyze page. The limit was increased to approximately 300,000 records.
Floe Page
Floe Reports no longer use v1 collections.
A menu has been added to the cube group to provide the group metrics.
The floe description in the Job Form is now collapsible.
The complete path for a job is now visible in its Job Details on the Jobs tab.
There is now the capability to right-click on a job name to open it in a new tab.
Floe Editor Tab
Floes created in the Floe Editor can now be validated using the new “Validate Floe” button on the toolbar. This button is greyed out when running the floe.
Job information, including parameters, can now be viewed for deleted floes.
An error message is now displayed when invalid floe specifications are pasted.
System Page
The “Details” option for Molecule Search databases includes a drop-down to select a date range to view logs.
Bug Fixes
Inappropriate caching caused a 500 error and required emptying the browser cache before Orion would load. This issue has been fixed.
An issue has been fixed that caused the plot to break when toggling between datasets with unrelated columns.
An issue has been fixed where a floe would fail if the directory name was too long. Now a floe cannot start if the directory name is longer than 86 characters. Folder names in the Data page are also limited to 86 characters.
The 2D depiction in the corner of the 3D Viewer lagged after a new record was selected to overlay search results versus a 3D query, which made triaging a hit list difficult. This issue has been fixed.
An issue has been fixed where molecules with certain types of binary data could cause the 3D Viewer to freeze.
An issue has been fixed where CIF files were not automatically converted to datasets using the direct import feature.
Orion Back End Notes
Note
Orion back end and other Orion changes of interest to all users are listed in this section, in this manual. Orion Platform changes of interest to Python programmers are summarized in the Orion Programming Guide release notes.
When a whole CPU/GPU is requested, an entire CPU/GPU is now assigned instead of available portions of multiple processors.
Spot termination no longer contributes to parallel cube throttling.
Information for Programmers
Note
The Programming Guide covers only Orion Platform.
Managed Service customers can use the Orion Platform interfaces to develop their own cubes and floes for execution in Orion. They can use the Orion Command-Line Interface (OCLI) to execute operations using the Orion Client facilities. They can also develop and manage Molecule Search databases. For more information, see the Programming Guide.