VAAST by KBase

VAAST is a web application for exploring genetic data and taxonomy using interactive visualizations and an LLM-powered chatbot interface.

Installation

1. Clone or Pull the Repository Ensure you have the latest version of the code:

   git pull origin main

2. Install Dependencies This project uses Poetry for dependency management. Install the dependencies with:

   poetry install

Data Setup

The application requires specific data files that are too large to be stored in version control.

1. Download Data Files You need to obtain the following pickle (.pkl) files from FigShare:

   • anthropic-docs-all.pkl
   • openai-docs-all.pkl

2. Place Files Move these two files into the data/ directory within the project root:

   vaast-app/
   └── data/
       ├── anthropic-docs-all.pkl
       └── openai-docs-all.pkl
   

Starting the App

To start the application, rely on Poetry to handle the environment:

poetry run python -m vaast_app.app

Alternatively, if you have activated the poetry shell (poetry shell), you can run:

python -m vaast_app.app

The application will start on http://0.0.0.0:8000 (accessible via http://localhost:8000).

Configuration

You can configure the application using environment variables or through the Settings panel in the user interface.

Environment Variables

The following environment variables can be set:

• OPENAI_API_KEY: API key for OpenAI (required if using OpenAI models).
• ANTHROPIC_API_KEY: API key for Anthropic (required if using Anthropic models).
• VERSION: Specifies the backend provider version (e.g., OpenAI, Anthropic).
• HOSTING_LOCATION: Specifies the hosting location (e.g., API, CBORG).
• MODEL: Specifies the model to use.

You can set these in your shell before running the application:

export OPENAI_API_KEY="your-api-key"
export VERSION="OpenAI"
poetry run python -m vaast_app.app

Settings Panel

Alternatively, you can configure these settings within the application:

1. Start the app and open it in your browser.
2. Click the Settings button in the top right corner.
3. Enter the values for the relevant keys and configuration options.
4. Click Save. The application will potentially reload updated documentation data if the Version is changed.

Using the App

Once the application is running, open your web browser and navigate to: http://localhost:8000

The application consists of:

• Landing Page: The entry point for the application.
• Visualizer: Explore genetic data and taxonomy trees interactively.
• Chatbot: Use the integrated chatbot to ask questions about the data, powered by LLM contexts loaded from the data files.