Skip to content

DELTA.md

Latest commit

 

History

History
92 lines (68 loc) · 1.79 KB

DELTA.md

File metadata and controls

92 lines (68 loc) · 1.79 KB
Raw

DELTA

Category: Chemistry
Repository: https://github.com/deltawave-tech/delta

Overview

DELTA is a multi-agent drug design pipeline that coordinates structure-based molecular docking (AutoDock), protein-ligand interaction analysis (PLIP), and LLM-driven design cycles. Three external services must be running before the pipeline starts.

Requirements

  • Python 3.x with uv package manager
  • Haskell cabal build tool (for Gurnemanz service)
  • Docker (for AutoDock service)
  • OpenAI or Anthropic API key

Installation

1. Install uv

curl -LsSf https://astral.sh/uv/install.sh | sh
# or
pip install uv

2. Clone and Install Dependencies

git clone https://github.com/deltawave-tech/delta.git
cd delta
uv sync
source .venv/bin/activate

Alternative (pip):

python3 -m venv venv
source venv/bin/activate
pip install --upgrade pip
pip install -e ".[dev]"

3. Configure API Keys

# Create .env in project root
cat > .env <<EOF
OPENAI_API_KEY="your-openai-key"
ANTHROPIC_API_KEY="your-anthropic-key"
EOF

Starting Required Services

All three services must be running before executing the pipeline:

1. Gurnemanz service:

cd gurnemanz
cabal build
cabal run gurnemanz

2. PLIP API service:

uv run python plip/plip/plip_api.py

3. AutoDock service:

cd autodock
docker compose up --build

Usage

uv run python src/demo/multi_agent_pipeline.py \
  --llm_provider sonnet-4 \
  --iteration_num 3

Options

Flag Default Description
--llm_provider sonnet-4 sonnet-4, sonnet-3.7, o3, gpt4.1, gemini
--iteration_num 3 Number of design/docking cycles

Output is saved under the runs/ directory with conversation logs and results.