DNA, RNA, and AA sequences often contain multiple different parts and relate to one another. For example, a plasmid might have a backbone, promoter, gene of interest, and a terminator. It may also produce an RNA sequence and/or a protein. In Benchling, these relationships are tracked in Registry using "biointelligent links." Specifically, Benchling supports three types of sequence links: part links, transcription links, and translation links. Biointelligent links allow scientists to define and track sequence and sub-sequence relationships.
Biointelligent link overview
Biointelligent links enable users to find, validate, and navigate relationships between sequences. These relationships are represented both visually on the sequence map and structurally within schema fields, ensuring consistency between what users see and the underlying data model.
Benchling offers multiple biointelligent link types:
-
Part link:
- DNA sub-sequences that are components of a larger DNA sequence
- RNA sub-sequences that are components of a larger RNA sequence
- Transcription link: An RNA sequence transcribed from a corresponding DNA region
- Translation link: A protein/AA sequence encoding for a portion of a DNA or RNA sequence
Here is a summary of the available link types that can be configured as metadata fields on each sequence type:
| Sequence Type | Part Link | Transcription Link | Translation Link |
| DNA sequence | ✓ | ✓ | ✓ |
| RNA sequence | ✓ | X | ✓ |
| AA sequence | X | X | X |
Configuring biointelligent links
Part, transcription, and translation links are configured using schema fields. To use these links, first create an entity schema of the appropriate type (DNA, RNA, or amino acid) to register sequences in Benchling. For more information, see the schema creation help article.
After creating the schema, you can add biointelligent links to your sequences. The available link types depend on the sequence type:
- DNA sequences support part, transcription, and translation links
- RNA sequences support part and translation links
To configure these links:
- Click your Avatar in the lower left corner
- Hover over Feature settings
- Select Registry settings
- Navigate to the entity schema you would like to add the biointelligent link field to
- In the Entity fields section, click the blue plus or the Add field button
- Give your field a Name
- Open the Definition dropdown
- Hover over the desired biointelligent link type - Part, Translation, Transcription - and select the desired biointelligent link entity schema
Configuring autofill rules
Autofill rules allow admins to configure how Benchling should automatically populate biointelligent links. Admins can configure how many links are added, if the links can overlap, the min and max length of the linked sequences, and more.
Note: If you do not see the ability to configure autofill rules on your tenant, please reach out to Benchling support. To use autofill rules, you must have the current part system enabled. See more details on our part systems below
To set up autofill rules:
- Navigate to the entity schema with biointelligent links configured
- In the Entity fields section, hover over the field you would like to configure the rule on and click the gear icon on the right hand side
- Select Set autofill and validation rules
- Configure your autofill rule in the pop-up modal
Available autofill rules and configurations
Autofill rules are configured on individual schema fields. For example, Field 1 might have overlapping parts disabled. This rule would not apply to any other biointelligent link fields on the entity schema.
Below you will find a summary of the currently available autofill rules that can be configured:
| Rule | Available Biointelligent Link Type(s) | Description | Additional Notes |
| Disable overlapping parts | Part links | If a schema field has more than one part that can be linked onto the sequence and these parts overlap one another, Benchling will only link out the longest part | This rule is configured independently for each part link field and is not applied globally. For example, if overlapping parts are disabled for both Field 1 and Field 2, Benchling will prevent overlaps within each field individually. However, parts linked in Field 1 may still overlap with parts linked in Field 2. |
| Only fill longest [link] | Part, transcription, and translation links | Only the longest part(s) up to the defined limit will be autofilled | As an example, if this field is set to 5, only the 5 longest parts are linked. If a new matching part is added that is longer than an existing linked part and autofill is rerun, the new part is added, resulting in 6 linked parts. The shortest part is then flagged as a validation error and can be removed to resolve the issue. |
| Limit base/amino acid lengths | Part, transcription, and translation links | Allows configuring the minimum and/or maximum length a biointelligent link sequence must be to be autofilled | If a part is manually added to a sequence and doesn’t meet the configured minimum and/or maximum length rule, it will be flagged as a validation error. |
| Require value | Part, transcription, and translation links | Schema field must be filled in prior to entity registration | We do not recommend configuring this on biointelligent link fields as these fields are not able to be filled out in a registration table. |
| Allow multi-select | Part, transcription, and translation links | More than one biointelligent link can be linked to the sequence in this field | Biointelligent links work best as multi-select fields. If autofill finds more than one linked sequence and the link field is single-select, it will error. |
Autofilling biointelligent links
Biointelligent links can be automatically populated in an entity’s metadata field and on the sequence map using the autofill functionality. The autofill workflow will be the same regardless of if you have configured autofill rules on the entity schema or not.
Autofill biointelligent links for a single entity
To autofill biointelligent links on one entity:
- Navigate to the Metadata tab of sequence you would like to autofill on
- Click the Autofill button in the upper right corner of the schema information
- If there are more than one type of biointelligent links configured, select which link type you would like to autofill from the dropdown
Note: Autofill translation uses standard genetic codes to find translations. However, if there is already a translation on the sequence with the same genetic code, you can link a translation from an alternative genetic code.
Once a biointelligent link has been added, the schema field will populate with entity chips
Additionally, an annotation will appear on the sequence map. How the annotation appears on the sequence map varies based on the type of biointelligent link. You can see what each annotation will look like in the screenshots below:
Bulk autofill biointelligent links
Biointelligent links can be autofilled on multiple entities at once by:
- Navigating to Global Search
- Filtering for the entities you would like to autofill and selecting them
- Click the More dropdown in the upper right corner, hover over Analyze, and select the appropriate Autofill [link] option from the dropdown
Manually creating part links
Instead of using autofill, users have the option to create parts links directly on a DNA sequence map. This workflow will create a part on your sequence map and add the part to the containing sequence’s schema field. To manually create a part link:
- Highlight the portion of the sequence representing the part link
- Right click on the highlighted region
- Select Create new part from the dropdown menu
- Give the part a name and select the schema field for the part link to be added to in the Replace selection with part pop-up modal.
- The schema type field will typically autopopulate based on the part link schema field configuration. However, if there is more than one schema type possible for your part link, you will need to verify that the right one is selected.
Note: Once you have completed this workflow, the part link still needs to be added to the registry. Otherwise, upon registering your containing sequence, you will receive a validation error. To do so, navigate to the part link entity and click the Register button in the upper right corner of the Metadata tab
Note: If a part link is manually added and doesn’t comply with the autofill rule configured at the schema level, a validation error will appear on the schema field
Removing and unlinking biointelligent links
Removing and/or unlinking a biointelligent link breaks the lineage between the part, translation, or transcription link and the containing sequence. Biointelligent links must be removed from both the sequence and schema field to ensure they are properly unlinked.
Removing links from schema fields
Part, transcription, and translation links can all be removed directly from their schema fields. For transcription and translation links, this fully disassociates the link and containing sequence from one another. For part links, this does not fully disassociate the lineage between the part and containing sequence (see below for more details).
To remove biointelligent links from a schema field:
- Navigate to the Metadata tab of the entity
- In the Schema section, click the Edit button
- Remove the biointelligent link from the schema field by clicking the X in the entity chip
For each biointelligent link type, the effects of removing the entity from the schema field can differ:
- Part links: Removing parts from a schema field does not automatically unlink it from the sequence map. The part will need to be manually unlinked from the sequence map as described below to fully break the biointelligent linkage.
- Transcription links: After removal from the schema field, the biointelligent linkage is broken. However, the translation will remain as an annotation from the sequence map as a standard sequence feature.
- Translation links: Removing a translation from a schema field automatically removes it from the sequence map as well.
Unlinking/deleting biointelligent links from the sequence map
Part and translation links can be removed directly on the sequence map. However, this does not disassociate the two entities from one another. In addition to removing the link from the sequence map, the link will also need to be removed from the schema field as described above.
To remove part and translation links from the sequence map:
- Navigate to the entities sequence map and locate you part or translation
- Right click on the part or translation and select one of the following based on the link type:
- Unlink part
- Delete translation
Benchling's biointelligent link systems
Benchling currently supports two biointelligent link systems, one of which is being gradually deprecated. The primary differences between these systems relate to how part links are structured and managed, as outlined below.
Customers who onboarded with Benchling earlier are likely using the legacy parts system, while all new customers are provisioned on the updated parts system. Over time, all customers using legacy parts will be transitioned to the new system. Benchling will provide advance notice and guidance ahead of any required migration.
Comparison of part systems behavior
Benchling’s current part system
Benchling evolved our part system to provide greater administrative and data integrity controls. The new part system is auto-enabled on Benchling tenants created from January 22, 2026 on.
Note: Once new parts is enabled, you are not able to return to the legacy part system
Benchling’s legacy part system
Unless enablement of the new part system has been requested for your tenant, the legacy part system is the default for all tenants created before January 22, 2026.
The major differences in product behavior between the two part systems are described in the table below:
| Functionality | Description | Current parts | Legacy parts |
| Overlapping parts | Parts representing different features linked onto the same portion of a sequence | ✓ | X |
| Nested parts | See nested parts section below | X | ✓ |
| Configurable autofill rules | Rules that affect how biointelligent links are autofilled | ✓ | X |
| Parts crossing sequence origin | Parts that overlap the sequence index | ✓ | X |
| Editing parts linked on sequences | Ability to edit part sequences once they have been linked onto containing sequences | X | ✓ |
Nested parts
A nested part is a sequence contained in and inherited from another part, then linked on the sequence that inherited it. For example, Backbone 1 has Promoter 1 part linked onto it. When Backbone 1 is linked onto Plasmid A, Promoter 1 would also be linked as a nested part. When nested parts are present, the part link metadata field will have a small arrow next to the field name.
Upon enabling the new parts system, all nested parts with existing links on sequences will be marked with a validation error. These validation errors can be resolved by rerunning autofill on the affected sequences.