Changelog

Source: NEWS.md

hgnc.mcp 0.1.0 (Development Version)

Updates (2025-11-11)

MCP Protocol Update

• Updated to plumber2mcp v0.4.0 with MCP protocol version 2025-06-18
• Updated test scripts and CI workflows to use new protocol version
• Breaking change: Clients must support MCP protocol version 2025-06-18 or later
• HTTP transport now includes MCP-Protocol-Version header
• Enhanced protocol negotiation support

Major Features

Data Management

• Smart caching system for HGNC complete dataset with automatic updates
• Cross-platform cache directory management (Linux, macOS, Windows)
• Configurable cache freshness (default: 30 days)
• Cache metadata tracking (download time, file size, source URL)

HGNC REST API Client

• Robust REST API client with rate limiting (≤10 req/sec)
• Session-level caching for repeated queries using memoise
• Comprehensive error handling and retries with exponential backoff
• Support for all major HGNC API endpoints

Gene Lookup and Resolution

• hgnc_find() - Search across symbols, aliases, and names
• hgnc_fetch() - Retrieve complete gene records
• hgnc_resolve_symbol() - Resolve symbols with strict/lenient modes
• hgnc_xrefs() - Extract cross-references (Entrez, Ensembl, UniProt, OMIM)

Batch Operations

• hgnc_normalize_list() - Fast batch normalization using local cache
• build_symbol_index() - Create efficient in-memory lookup indexes
• Support for handling ambiguous matches and withdrawn genes

Gene Groups and Families

• hgnc_group_members() - Get all genes in a specific group
• hgnc_search_groups() - Discover groups by keyword search

Change Tracking and Validation

• hgnc_changes() - Track nomenclature changes by date
• hgnc_validate_panel() - Validate gene panels against HGNC policy
• Comprehensive validation with replacement suggestions

MCP Server

• Model Context Protocol (MCP) server for AI assistant integration
• 10 MCP tools for gene nomenclature operations
• 4 MCP resources for context injection (gene cards, group cards, etc.)
• 4 MCP prompts for guided multi-step workflows
• Interactive Swagger UI for API documentation
• Standalone server script with command-line configuration

Docker Deployment

• Production-ready Docker image with multi-stage builds
• Docker Compose configuration with health checks
• Multi-platform builds (amd64, arm64)
• Published to GitHub Container Registry

Documentation

• 4 comprehensive vignettes:
  • Getting Started with hgnc.mcp
  • Normalizing Gene Lists for Clinical Panels
  • Working with HGNC Gene Groups
  • Running the MCP Server
• Complete function documentation with examples
• Real-world use cases and workflow examples

Testing and Quality

• Comprehensive test suite with >80% coverage
• Unit tests for all major functions
• Integration tests for REST API and caching
• Automated testing across multiple platforms (Linux, macOS, Windows)
• Automated Docker build and test pipeline

Initial Release

This is the first development release of hgnc.mcp. The package provides a complete toolkit for working with HGNC gene nomenclature in R, with particular focus on:

1. Speed: Local caching for fast bulk operations
2. Reliability: Robust error handling and rate limiting
3. Usability: Clean API with sensible defaults
4. Integration: MCP server for AI assistant workflows
5. Reproducibility: Version tracking and provenance

The package is ready for use in research and production environments, with active development continuing for additional features and improvements.