Insights Labs is a new beta program at Benchling for the Insights area. In order to get new charts types, chart customizations, and other UI upgrades in your hands as fast as possible, we've created this voluntary program.

New features will be added rapidly to this beta program and clearly labeled as beta. Most features added to Insights Labs will be available there for a few weeks or months before being broadly released. Keep in mind that though we will do our utmost to avoid it, features in this set may break, disappear, or change without warning. For more about what it means for features to be in beta, see our article that describes the Benchling release stages. 

Please keep in mind that an analysis or analysis template that is created with an Insights Labs feature may but is not guaranteed to work for a user who does not have Insights Labs enabled.

Getting Access

Your tenant administrator can enable access to "Insights Labs" by navigating to the Tenant admin console > Settings  > Insights. This can be enabled for specific users, all users. The default is disabled for all users.

Current Feature List

Experiment Optimization 

Experiment Optimization is a powerful analysis tool that helps you understand and optimize your experiments. It's designed to solve complex, multi-dimensional problems where an outcome is affected by many interacting variables. For more information on how Experiment Optimization works, see the linked article. 

Note: To access this feature in beta you must have a license to either Advanced Analysis or Bioprocess. 

Create Tables from SQL Queries in Analysis

You can now create tables in Analysis by writing a direct SQL query against the Benchling Warehouse. This gives you a flexible, code-first way to pull precisely the data you need into an Analysis without relying solely on dashboard blocks or registry imports. You can also parameterize a query so it draws values from an upstream dataset, which lets one query adapt to whatever data flows into it.

Create a SQL query table

1. In the Analysis flowchart, click Add source step
2. Under Create table, select Create table from SQL query 
3. Click Next 
4. Enter a Table name 
5. Write a valid SQL query against the Benchling Warehouse in the SQL query editor 
6. Click Show preview to check your results before committing 
7. Click Save

The resulting table is created in your Analysis just like any other dataset, and can be used as input to downstream transforms, charts, and other steps.

Use parameters in a SQL query

Parameters let a query pull values from an upstream dataset instead of hard-coding them. You connect an input dataset, map each parameter to a column in that dataset, then reference the parameter in your query. This is useful when the values you want to filter on are themselves the output of an earlier step.

1. From an existing dataset in your Analysis, create a new Create table from SQL query step
2. Open the configuration sidebar for your SQL query step 
3. Under Parameters table, select the upstream table that contains the values you want to use 
4. Under Parameters, click + 
5. Enter a Parameter name — this is the name you reference in the query 
6. Under Input column, select the column the parameter draws its values from 
7. Under Multiple values, select Use all values (up to 10,000) to pass every value in the column. Leave this unselected to pass only the value from the first row 
8. Reference the parameter in your query using double curly braces, for example {{min_value}} 
9. Click Save

Data limits:

• Up to 50,000 rows can be returned by a SQL query step — consistent with the existing "Create from dashboard block" and CSV export limits.
• A multi-value parameter passes up to 10,000 values

Current limitations:

• Features like a schema browser, SQL Writer are planned for a future iteration.

• When an Analysis containing a SQL query step is saved as a template, the SQL table is treated as Static — the query does not re-execute when the template is applied. Users will need to manually create or update the table when starting from that template.

  • This does not apply to editable Automation templates (limited availability). A Create table from SQL query step added directly to an editable Automation template re-executes every time the template is applied, so the table reflects the data available at apply time rather than data frozen at template creation.

• In all cases, queries run against the Data Warehouse, so results reflect the most recently synced data rather than a live, real-time view.

Visual Flagging

Visual flagging tool allows you to use charts and graphs to identify and mark data points that appear to be significantly different from the rest of the data. This tool allows you to flag the unusual data points for further investigation or exclusion.

You can apply visual flagging to the following plots: 

• Scatter Plot
• Regression -- Note: Regressions are only available to users with access to Advanced Analysis

For more information on how to use the feature visit the Visual Flagging section of our Analyze Data with Benchling Analysis article.