Provides access to OpenRouter's NVIDIA Nemotron 3 Ultra model for variant consequence and clinical significance queries.
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@VepClin-MCP annotate variant rs80357413"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
VepClin-MCP is a terminal-based bioinformatics CLI chat tool that integrates Ensembl VEP, NCBI ClinVar, a custom-built MCP server layer, and OpenRouter's NVIDIA Nemotron 3 Ultra model to look up variant consequences and clinical significance, presenting the results as clear, readable summaries in a Rich-powered CLI.
NVIDIA Nemotron 3 Ultra powered chat interface
Ensembl VEP integration for genomic and transcript-qualified HGVS variant consequence lookup
ClinVar integration for clinical significance, oncogenicity, review status, traits, & variation IDs
Export batch results as CSV, TSV, annotated VCF, or multi-sheet Excel (.xlsx)
Single-variant PDF report generation
Gruvbox-styled Rich terminal interface with readable panels, tables, & status messages, featuring pixel-art mascot Exon the Axolotl
MCP server layer exposing custom reusable variant annotation tools
Related MCP server: MCP Chat
Command | Description |
| Upload VCF files & summarize multiple variants (up to 200 at a time) |
| Save latest batch results as CSV, TSV, VCF, or Excel |
| Save single-variant lookups as a PDF report |
| Switch between GRCh38 and GRCh37 lookups |
| Choose MANE Select-only results or all transcript consequences |
| Reset the conversation context |
| Show usage tips and this command list |
| Quit VepClin |
CLI/UI: Python, Rich, Questionary
MCP Layer: FastMCP
HTTP/API Client: httpx
AI: OpenRouter API, NVIDIA Nemotron 3 Ultra
Variant Annotation: Ensembl VEP REST API
Clinical Data: NCBI ClinVar via Biopython Entrez
Excel/PDF Export: openpyxl, ReportLab
Storage: Local config.json file for genome build & transcript-mode preferences
Packaging: setuptools, pyproject.toml
Testing: pytest, FastMCP test client
---
---
config:
layout: dagre
theme: redux-dark-color
look: redux
fontFamily: '''Source Code Pro Variable'', monospace'
themeVariables:
fontFamily: '''Source Code Pro Variable'', monospace'
---
flowchart LR
User[User] --> CLI[VepClin CLI<br/>terminal_ui.py<br/>Rich + Questionary]
CLI --> Runner[Chat + Command Loop<br/>llm_runner.py<br/>Handles chat, slash commands, tool calls]
Runner --> MCP[Embedded FastMCP Server<br/>mcp_server.py<br/>Exposes variant tools]
MCP --> Client[VariantClient<br/>variant_client.py<br/>Runs VEP + ClinVar lookups]
Client --> VEP[Ensembl VEP REST API<br/>Variant consequences]
Client --> ClinVar[NCBI ClinVar / Entrez<br/>Clinical significance]
VEP --> Results[Annotated Variant Results<br/>VEP consequence data]
ClinVar --> Results[Annotated Variant Results<br/>ClinVar clinical data]
Results --> Runner
Runner --> CLI
Runner --> Settings[Session Settings<br/>session_config.py<br/>Genome build + transcript mode]
Runner --> Exports[Export Helpers<br/>exporter.py<br/>CSV / TSV / VCF / XLSX / PDF]
Exports --> Files[Local Output Files<br/>Reports + batch exports]python -m venv .venv
.venv\Scripts\Activate.ps1
python -m pip install git+https://github.com/shivankvirdi/VepClin-MCP.gitpython3 -m venv .venv
source .venv/bin/activate
python -m pip install git+https://github.com/shivankvirdi/VepClin-MCP.gitOPENROUTER_API_KEY is required for the LLM chat interface (https://openrouter.ai/).NCBI_EMAIL is recommended for consistent NCBI Entrez/ClinVar requests.NCBI_API_KEY is optional, but recommended for higher ClinVar request limits (https://www.ncbi.nlm.nih.gov/datasets/docs/v2/api/api-keys/).
[Environment]::SetEnvironmentVariable("OPENROUTER_API_KEY", "sk-or...", "User")
[Environment]::SetEnvironmentVariable("NCBI_EMAIL", "you@example.com", "User")
[Environment]::SetEnvironmentVariable("NCBI_API_KEY", "...", "User")# Zsh
echo 'export OPENROUTER_API_KEY="sk-or..."' >> ~/.zshrc
echo 'export NCBI_EMAIL="you@example.com"' >> ~/.zshrc
echo 'export NCBI_API_KEY="..."' >> ~/.zshrc
source ~/.zshrc# Bash
echo 'export OPENROUTER_API_KEY="sk-or..."' >> ~/.bashrc
echo 'export NCBI_EMAIL="you@example.com"' >> ~/.bashrc
echo 'export NCBI_API_KEY="..."' >> ~/.bashrc
source ~/.bashrcgit clone https://github.com/shivankvirdi/VepClin-MCP.git
cd VepClin-MCP
python -m venv .venv
.venv\Scripts\Activate.ps1
python -m pip install -e .
Copy-Item .env.example .envgit clone https://github.com/shivankvirdi/VepClin-MCP.git
cd VepClin-MCP
python3 -m venv .venv
source .venv/bin/activate
python -m pip install -e .
cp .env.example .envFollow .env.example and add your API keys & email to VepClin-MCP/.env in the repo root.
Run the terminal chat CLI:
vepclinMost users don't need this. The vepclin chat CLI starts and uses the MCP tools automatically.
vepclin-serverIf you installed from source, you can also run the server directly:
python backend/mcp_server.pyThis server cannot be installed
Unclaimed servers have limited discoverability.
If you are the server author, to access and configure the admin panel.
MCP gateway federating 21 biomedical MCP servers behind one endpoint: gnomAD, ClinVar, HPO, VEP.
Real-time chat hub for AI agents — Claude Code, Cursor, Cline, Codex over MCP or REST.
Real-time chat hub for AI agents — Claude Code, Cursor, Cline, Codex over MCP or REST.
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/shivankvirdi/VepClin-MCP'
If you have feedback or need assistance with the MCP directory API, please join our Discord server