Why use the Registry to validate data?
Benchling's Registry acts as data repository for an organization's scientific entities. It is used to track these entities through the entire research process and across several groups. Since this data may be shared across several internal groups, it must be kept error-free in order to make sure the most accurate scientific conclusions are made.
Benchling can validate your organization's registry to ensure data integrity.
Leverage unique constraints
You can leverage Benchling's uniqueness constraint feature to ensure that only unique DNA or AA sequences are registered. Further configurations can also check if the metadata (ie. schema links, part links, and etc.) are filled out correctly according to the uniqueness constraints for each schema.
Registry admins can change these uniqueness constraints within their organization's Registry settings:
If entities within an import violate any configured uniqueness constraints, registration will fail until the violation is resolved.
Automatically invalidate a registered identity
After registration, the metadata of these entities can be modified by users - potentially by mistake. These errors can cause unseen errors that can affect downstream experiments.
Use cases that lead to entity invalidation include:
-
Editing fields such that two entities contain exactly similar metadata.
-
Removing metadata that was initially required for registration.
-
Archiving an entity/part that was linked to a required field.
To identify these errors, Benchling will perform weekly validation checks on the registry, split deterministically across Friday, Saturday, and Sunday when running these comprehensive checks has the least impact on end user experience. If Benchling perceives that entities are invalid, it will mark those entities and provide information about their invalidation.
Interpret notifications about invalid entries
Benchling will notify users of registered entities that fail validation:
-
A user will receive an email notification if they invalidate an entity by modifying its metadata.
-
Invalidated entities will have a red "Failed" tag next to it.
-
Clicking into the "Failed" entity will provide more information on what is causing the failed validation check and how to fix said errors. Click into the "?" icon to get more information about the error type.
-
Users can screen for invalidated entities by navigating to Benchling's search bar and filtering based on "Validation Status".
-
Users can also further investigate invalidated entities by filtering based on "Failure Reason":
-
Invalidated entities that are @-mentioned will be indicated with a red chip, however it does NOT stop a user from utilizing it - users can still click the chip to get the entity information.
-
Invalidated entities can be manually overridden only by registry admins. However all organization members will be able to see that the validation check was overridden. Note: These checks can be overriden in bulk.
-
To resolve uniqueness conflicts, unregister one of the duplicates and re-register it to merge it into the other entity that remained registered.
Example: "Plasmid A" and "Plasmid B" were registered in the Registry, and a user changes one of the fields in the plasmids such that they both had similar metadata. Benchling will invalidate these entities. If a user checks his notes, unregisters Plasmid A, and re-registers it, "Plasmid A" will be merged to "Plasmid B". All notes and linkages associated with "Plasmid A" will be transferred to "Plasmid B".