Use Registry and Result schema data to create a dataset. Visit Creating an analysis for information on creating a new analysis.
Note: This feature is currently in limited availability. If you don’t have access to the new Analysis feature, contact Customer Support to request access.
Create a dataset
- Click Add dataset in the main view.
- Select the “From Registry schema” option in the menu.
- Select a Registry schema to start from. This will pull all entities from the selected schema into the dataset.
- Apply filters to narrow down the entities that are pulled into the dataset.
- Select the columns to include in the dataset.
- Entity links on the schema can be expanded to allow linked schema data to be pulled into the dataset.
- Click Add Results to add Results that are recorded against entities in the selected Registry schema.
- If there is more than one field on the Result schema that could link to the Registry schema, a dropdown will appear to specify which field of the Result schema that you want to join on.
- Selecting Only include latest results will pull the latest result captured per entity into the dataset, whereas Include all results will include all result data captured against the entity.
- Up to 50,000 rows of data can be created.
- This method of dataset creation queries Benchling’s core database, and therefore does not have any sync latency.
- Note that when multi-select fields or related results with >1 value are selected, the data is split across multiple rows. Linked data is joined via a “left join”. To re-aggregate data in downstream analyses, we recommend selecting the “id” column when needed.
- The created dataset is a snapshot of the data returned from the executed query. The dataset does not update over time.