sparql-llm

Overview

SPARQL-LLM is a comprehensive solution that combines an MCP server, chat system, and reusable components to enhance LLM capabilities when generating SPARQL queries for biological databases. It exposes tools for retrieving relevant documents, class schemas, and executing SPARQL queries against endpoints like UniProt and Bgee. The system integrates Retrieval-Augmented Generation with SPARQL query validation through endpoint schemas to ensure accuracy in query generation on large-scale knowledge graphs.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

• retrieve_relevant_documents

  Retrieve relevant documents to help writing SPARQL queries for SIB biodata resources

  Try:Find me examples of SPARQL queries for protein informationTry:Show me documents about gene expression patterns in different species
• retrieve_relevant_classes_schema

  Retrieve relevant class schema to help writing SPARQL queries for SIB biodata resources

  Try:Get me the class schema for protein annotationsTry:Show me the available classes in the UniProt endpoint
• execute_sparql_query

  Execute a SPARQL query against a specified endpoint

  Try:Find all proteins from human that have 'kinase' in their nameTry:Get all metabolic pathways involving lipids from SwissLipids

Comparable tools

Installation

pip install sparql-llm

uv add sparql-llm

{
  "servers": {
    "expasy-mcp": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "sparql-llm"
      ]
    }
  }
}

{
	"servers": {
		"expasy-mcp-http": {
			"url": "https://chat.expasy.org/mcp/",
			"type": "http"
		}
	}
}

FAQ

What biological endpoints does this MCP server support?

The server supports multiple SIB biodata resources including UniProt, Bgee, OMA, SwissLipids, and Cellosaurus through their SPARQL endpoints.

How do I validate my SPARQL queries with this system?

You can use the validation function that checks if predicates/types used are compliant with the VoID description in your target SPARQL endpoint. It supports federated queries and path patterns.

Compare sparql-llm with