Tonina - Multi-Agent Bioinformatics Analysis System

Tonina is a comprehensive multi-agent bioinformatics analysis system that extends the capabilities of gffutilsAI to handle multiple types of biological data through specialized AI agents.

Features

Multi-Agent Architecture

• GFF Agent: Complete GFF file analysis (ported from gffutilsAI)
• Sequence Agent: FASTA sequence analysis and manipulation
• Proteomics Agent: Protein structure and functional analysis
• Phylogenetics Agent: Evolutionary analysis and tree operations
• General Agent: Cross-format utilities and file operations

Intelligent Query Routing

Tonina automatically routes your queries to the appropriate specialized agent based on:

• File types (.gff, .fasta, .pdb, etc.)
• Query content and keywords
• Domain-specific terminology

Multi-LLM Support

• Local Models: Ollama (llama3.1, codellama, etc.)
• Cloud Models: OpenAI GPT-4, Anthropic Claude, Google Gemini
• Flexible Configuration: Easy switching between providers

Installation

# Clone the repository
git clone <repository-url>
cd tonina

# Install dependencies
pip install -e .

# Or using uv (recommended)
uv pip install -e .

Configuration

1. Copy the example environment file:

cp .env.example .env

2. Edit .env with your API keys:

# For cloud models (optional)
OPENAI_API_KEY=your_key_here
ANTHROPIC_API_KEY=your_key_here
GOOGLE_API_KEY=your_key_here

Usage

Basic Commands

# Interactive mode with local model
tonina --model llama3.1 --server local

# Single query
tonina --query "What genes are in chromosome 1?" --model gpt-4o

# Batch processing
tonina --batch queries.txt --model claude-3-5-sonnet-latest

# Specific provider
tonina --anthropic --model claude-3-5-sonnet-latest
tonina --gemini --model gemini-2.0-flash-exp
tonina --openai --model gpt-4o

Agent-Specific Examples

GFF Analysis

tonina "Find all genes in chromosome 1 between positions 1000-5000"
tonina "Calculate feature statistics for my GFF file"
tonina "Export all membrane proteins to CSV"

Sequence Analysis

tonina "Analyze this FASTA file for ORFs"
tonina "Calculate GC content for all sequences"
tonina "Find motifs in this protein sequence"

Multi-Agent Workflows

tonina "Extract genes from GFF, get their sequences, and analyze domains"
# This automatically coordinates: GFF Agent → Sequence Agent → Proteomics Agent

Agent Capabilities

GFF Agent (31 tools)

• Coordinate-based queries and region analysis
• Feature statistics and length distributions
• Attribute searches and GO function analysis
• Intergenic region identification
• Comprehensive export capabilities

Sequence Agent (15+ tools)

• FASTA file parsing and validation
• Sequence composition analysis
• ORF finding and translation
• Motif searching and pattern matching
• Multiple sequence alignment

Proteomics Agent (15+ tools)

• Protein structure analysis (PDB files)
• Domain prediction and annotation
• Functional classification
• Comparative protein analysis

Phylogenetics Agent (10+ tools)

• Phylogenetic tree parsing
• Evolutionary distance calculations
• Species comparison
• Ortholog identification

General Agent (10+ tools)

• Cross-format file operations
• Data integration workflows
• Format conversion utilities
• Batch processing tools

Development

Adding New Agents

1. Create agent directory: src/tonina/agents/new_agent/
2. Implement tools with @tool decorator
3. Create agent class extending BaseAgent
4. Register in AgentManager

Adding New Tools

from strands import tool

@tool
def my_analysis_tool(input_file: str, parameter: str) -> dict:
    \"\"\"Tool description for AI understanding\"\"\"
    # Implementation
    return {"result": "analysis_output"}

License

GPL-3.0 License - see LICENSE file for details.

Contributing

1. Fork the repository
2. Create a feature branch
3. Add tests for new functionality
4. Submit a pull request

Support

For issues and questions, please use the GitHub issue tracker.