Graphical User Interface

Configure your environment

Some of the options for Lettuce can be configured through the UI. Others, like database credentials are still configured through the environment. If you have changes to the defaults that work for the quickstart, then copy your .env file over.

Navigate to the /lettuce-ui directory

Run the UI with uv

uv run lettuce-ui

This will start the user interface in your browser

Configure your UI

To use the UI, some set-up is required.

First, you have to choose which search types to enable. There’s a sidebar in the app that allows you to enable different pipelines.

This lets you choose:

1. One of the pre-configured embeddings models by its short name
2. Whether the embeddings code is enabled
3. Whether the LLM is enabled
4. An Ollama URL if your environment is configured to use Ollama (default)
5. The model name to request from Ollama

Upload a CSV file

The UI assumes you have a CSV file with your source terms held in a column.

There’s a file upload button which will open a file picker, which you can use to upload your CSV.

Then, you can choose which column of the CSV contains source terms. Once you’ve selected this, the UI automatically runs keyword search on your source terms to suggest concepts.

(Optional) Select default domains and vocabularies

If you have pre-defined domains and vocabularies that you want to map all or most of your source terms to, you can set defaults in this top menu. You can choose nothing to search all domains and vocabularies by default.

Get suggestions

Once you have chosen a column of your CSV for your source terms, a table below the options will populate.

Perhaps you’re lucky and you get exactly what you want for each source term from the initial search. In most cases, you’ll want to change how Lettuce searches for concepts. For each source term you have options

Source term: An editable term to use for search. If you have an obvious typo in your source term, you can simply search for something else

Domain: A multi-select box for domains to search.

Vocabulary: A multi-select box for vocabularies to search.

Standard concept: A checkbox to only look for standard concepts

Valid concept: A checkbox to only look for concepts that haven’t been invalidated

Search mode: A select box to choose which kind of search to use

Suggestion: You may get multiple options to look at. This box is only really meant for you to check your suggestions quickly so you can see if you want to change your search parameters

Concept ID: If you have selected a suggestion, this shows you a link to Athena for this concept

Choose suggestions

Above the search table, there’s a tab selector. If you click “View suggestions”, you are shown more details about the concepts suggested.

Save selections

Once you’ve chosen suggestions you’re happy with, you can save them. You choose a filename and click save to create a CSV file with the following fields

Search parameters

• source_term
• search_term
• domains
• vocabs
• search_standard_concept
• valid_concept
• search_mode

Concept parameters

• concept_id
• concept_name
• domain_id
• vocabulary_id
• standard_concept
• score