enrichr-mcp-server
Version:
MCP Server for Enrichr gene set enrichment analysis with multi-library support
717 lines (710 loc) ā¢ 33.8 kB
JavaScript
/**
* MCP Server for Enrichr GO Biological Process Enrichment Analysis
*
* This server provides tools to query Enrichr for GO enrichment analysis,
* returning only statistically significant results (p < 0.05) to reduce context usage.
*/
import { Server } from "@modelcontextprotocol/sdk/server/index.js";
import { StdioServerTransport } from "@modelcontextprotocol/sdk/server/stdio.js";
import { CallToolRequestSchema, ListToolsRequestSchema, } from "@modelcontextprotocol/sdk/types.js";
import fetch from "node-fetch";
import { FormData } from "node-fetch";
import { writeFileSync } from "fs";
import { libraryDescriptions } from "./library_descriptions.js";
const ENRICHR_URL = "https://maayanlab.cloud/Enrichr";
const POPULAR_LIBRARIES = [
"GO_Biological_Process_2025",
"KEGG_2021_Human",
"Reactome_2022",
"MSigDB_Hallmark_2020",
"ChEA_2022",
"GWAS_Catalog_2023",
"Human_Phenotype_Ontology",
"STRING_Interactions_2023",
"DrugBank_2022",
"CellMarker_2024"
];
/**
* Parse command line arguments and environment variables for configuration
*/
function parseConfig() {
const args = process.argv.slice(2);
const config = {
defaultLibraries: POPULAR_LIBRARIES, // Default fallback
version: "0.2.1",
maxTermsPerLibrary: 50, // Default to 50 terms per library
format: "detailed", // Default to detailed format
saveToFile: false, // Default to not saving to file
outputFile: "" // Default output file path
};
// Parse CLI arguments
for (let i = 0; i < args.length; i++) {
const arg = args[i];
if (arg === '--libraries' || arg === '-l') {
const librariesArg = args[i + 1];
if (librariesArg) {
if (librariesArg.toLowerCase() === 'pop') {
config.defaultLibraries = POPULAR_LIBRARIES;
}
else {
config.defaultLibraries = librariesArg.split(',').map(lib => lib.trim());
}
i++; // Skip next argument since we consumed it
}
}
else if (arg === '--max-terms' || arg === '-m') {
const maxTermsArg = args[i + 1];
if (maxTermsArg) {
const maxTerms = parseInt(maxTermsArg);
if (!isNaN(maxTerms) && maxTerms > 0) {
config.maxTermsPerLibrary = maxTerms;
}
i++;
}
}
else if (arg === '--format' || arg === '-f') {
const formatArg = args[i + 1];
if (formatArg && ['detailed', 'compact', 'minimal'].includes(formatArg)) {
config.format = formatArg;
i++;
}
}
else if (arg === '--output' || arg === '-o') {
const outputArg = args[i + 1];
if (outputArg) {
config.outputFile = outputArg;
config.saveToFile = true;
i++;
}
}
else if (arg === '--compact' || arg === '-c') {
config.format = "compact";
}
else if (arg === '--minimal') {
config.format = "minimal";
}
else if (arg === '--help' || arg === '-h') {
console.log(`
Enrichr MCP Server
Usage: enrichr-mcp-server [options]
Options:
-l, --libraries <libs> Comma-separated list of default Enrichr libraries.
Use "pop" for a curated list of popular libraries.
(default: pop)
-m, --max-terms <num> Maximum terms to show per library (default: 50)
-f, --format <format> Output format: detailed, compact, minimal (default: detailed)
-o, --output <file> Save complete results to TSV file
-c, --compact Use compact format (same as --format compact)
--minimal Use minimal format (same as --format minimal)
-h, --help Show this help message
Format Options:
detailed - Full details: p-values, odds ratios, gene lists (default)
compact - Term name + p-value + gene count (saves ~50% context)
minimal - Just term name + p-value (saves ~80% context)
Environment Variables:
ENRICHR_LIBRARIES Comma-separated list of default libraries
ENRICHR_MAX_TERMS Maximum terms per library
ENRICHR_FORMAT Output format (detailed/compact/minimal)
ENRICHR_OUTPUT_FILE TSV output file path
Examples:
enrichr-mcp-server --libraries "GO_Biological_Process_2025,KEGG_2021_Human"
enrichr-mcp-server -l pop
enrichr-mcp-server -l "MSigDB_Hallmark_2020" --max-terms 20 --compact
enrichr-mcp-server --format minimal --max-terms 30 --output results.tsv
enrichr-mcp-server --minimal --max-terms 50 --output /tmp/enrichr_results.tsv
ENRICHR_LIBRARIES="GO_Biological_Process_2025,Reactome_2022" enrichr-mcp-server
Popular Libraries:
GO_Biological_Process_2025 - Gene Ontology Biological Processes
GO_Molecular_Function_2025 - Gene Ontology Molecular Functions
GO_Cellular_Component_2025 - Gene Ontology Cellular Components
KEGG_2021_Human - KEGG Pathways
Reactome_2022 - Reactome Pathways
MSigDB_Hallmark_2020 - MSigDB Hallmark Gene Sets
Human_Phenotype_Ontology - Human Phenotype Ontology
WikiPathways_2023_Human - WikiPathways
ChEA_2022 - ChIP-seq Experiments
GTEx_Tissue_Sample_Gene_Expression_Profiles_up - GTEx Tissue Expression
`);
process.exit(0);
}
}
// Override with environment variables if set
if (process.env.ENRICHR_LIBRARIES) {
config.defaultLibraries = process.env.ENRICHR_LIBRARIES.split(',').map(lib => lib.trim());
}
if (process.env.ENRICHR_MAX_TERMS) {
const maxTerms = parseInt(process.env.ENRICHR_MAX_TERMS);
if (!isNaN(maxTerms) && maxTerms > 0) {
config.maxTermsPerLibrary = maxTerms;
}
}
if (process.env.ENRICHR_FORMAT) {
const format = process.env.ENRICHR_FORMAT;
if (['detailed', 'compact', 'minimal'].includes(format)) {
config.format = format;
}
}
if (process.env.ENRICHR_OUTPUT_FILE) {
config.outputFile = process.env.ENRICHR_OUTPUT_FILE;
config.saveToFile = true;
}
return config;
}
// Parse configuration at startup
const CONFIG = parseConfig();
console.error(`š§¬ Enrichr MCP Server starting...`);
console.error(`š Default libraries: ${CONFIG.defaultLibraries.join(', ')}`);
console.error(`š Max terms per library: ${CONFIG.maxTermsPerLibrary}`);
console.error(`š Format: ${CONFIG.format}`);
if (CONFIG.saveToFile) {
console.error(`š¾ Output file: ${CONFIG.outputFile}`);
}
/**
* Query Enrichr for enrichment results from specified library/libraries
*/
async function queryEnrichrLibraries(geneList, libraries = ["GO_Biological_Process_2025"], description = "Gene list for enrichment") {
const genesStr = geneList.join("\n");
// 1. Add gene list to Enrichr
const addListUrl = `${ENRICHR_URL}/addList`;
const formData = new FormData();
formData.append('list', genesStr);
formData.append('description', description);
try {
const addResponse = await fetch(addListUrl, {
method: 'POST',
body: formData,
});
if (!addResponse.ok) {
const error = { error: `HTTP error ${addResponse.status}: ${addResponse.statusText}` };
return Object.fromEntries(libraries.map(lib => [lib, error]));
}
const addListData = await addResponse.json();
const userListId = addListData.userListId;
if (!userListId) {
const error = { error: `'userListId' not found in Enrichr's /addList response: ${JSON.stringify(addListData)}` };
return Object.fromEntries(libraries.map(lib => [lib, error]));
}
// 2. Get enrichment results for each library
const results = {};
for (const library of libraries) {
const enrichUrl = `${ENRICHR_URL}/enrich`;
const enrichParams = new URLSearchParams({
userListId: userListId.toString(),
backgroundType: library
});
try {
const enrichResponse = await fetch(`${enrichUrl}?${enrichParams}`);
if (!enrichResponse.ok) {
results[library] = { error: `HTTP error ${enrichResponse.status}: ${enrichResponse.statusText}` };
continue;
}
const enrichmentResults = await enrichResponse.json();
if (!(library in enrichmentResults)) {
results[library] = { error: `Library '${library}' not found in results. Available: ${Object.keys(enrichmentResults).join(', ')}` };
continue;
}
const allResults = enrichmentResults[library];
// Filter for significant results using adjusted p-value (adjusted p < 0.05)
const significantResults = allResults.filter((result) => result[6] < 0.05);
results[library] = {
total_terms: allResults.length,
significant_terms: significantResults.length,
results: significantResults
};
}
catch (error) {
results[library] = { error: `Error querying library ${library}: ${error instanceof Error ? error.message : String(error)}` };
}
}
return results;
}
catch (error) {
const errorResult = { error: `Error querying Enrichr: ${error instanceof Error ? error.message : String(error)}` };
return Object.fromEntries(libraries.map(lib => [lib, errorResult]));
}
}
/**
* Format enrichment results for multiple libraries
*/
function formatMultiLibraryResults(results, maxTerms = CONFIG.maxTermsPerLibrary, format = CONFIG.format) {
const outputLines = [];
for (const [library, result] of Object.entries(results)) {
outputLines.push(`\n=== ${library} ===`);
if ("error" in result) {
outputLines.push(`Error: ${result.error}`);
continue;
}
const { total_terms, significant_terms, results: libraryResults } = result;
if (significant_terms === 0) {
outputLines.push(`No significant terms found (adjusted p < 0.05) out of ${total_terms} total terms analyzed.`);
continue;
}
const actualTermsShown = Math.min(significant_terms, maxTerms);
outputLines.push(`Found ${significant_terms} significant terms (adjusted p < 0.05) out of ${total_terms} total terms:`);
if (significant_terms > maxTerms) {
outputLines.push(`Showing top ${actualTermsShown} terms (use --max-terms to show more):\n`);
}
else {
outputLines.push('');
}
if (format === "compact") {
// Compact format: more terms, less detail
libraryResults.slice(0, maxTerms).forEach((termInfo, i) => {
const [rank, termName, pValue, oddsRatio, combinedScore, overlappingGenes, adjustedPValue] = termInfo;
outputLines.push(`${i + 1}. ${termName} (adj-p: ${adjustedPValue.toExponential(2)}, genes: ${overlappingGenes.length})`);
});
}
else if (format === "minimal") {
// Minimal format: just term name and p-value
libraryResults.slice(0, maxTerms).forEach((termInfo, i) => {
const [rank, termName, pValue, oddsRatio, combinedScore, overlappingGenes, adjustedPValue] = termInfo;
outputLines.push(`${i + 1}. ${termName} (p: ${pValue.toExponential(2)})`);
});
}
else {
// Detailed format: fewer terms, more detail
libraryResults.slice(0, maxTerms).forEach((termInfo, i) => {
const [rank, termName, pValue, oddsRatio, combinedScore, overlappingGenes, adjustedPValue] = termInfo;
outputLines.push(`${i + 1}. ${termName}`);
outputLines.push(` Adjusted P-value: ${adjustedPValue.toExponential(2)}`);
outputLines.push(` Raw P-value: ${pValue.toExponential(2)}`);
outputLines.push(` Odds Ratio: ${oddsRatio.toFixed(2)}`);
outputLines.push(` Combined Score: ${combinedScore.toFixed(2)}`);
outputLines.push(` Overlapping Genes (${overlappingGenes.length}): ${overlappingGenes.join(', ')}`);
outputLines.push("");
});
}
}
return outputLines.join("\n");
}
/**
* Query Enrichr for GO Biological Process 2025 enrichment results
*/
async function queryEnrichrGoBp(geneList, description = "Gene list for GO BP enrichment") {
const genesStr = geneList.join("\n");
// 1. Add gene list to Enrichr
const addListUrl = `${ENRICHR_URL}/addList`;
const formData = new FormData();
formData.append('list', genesStr);
formData.append('description', description);
try {
const addResponse = await fetch(addListUrl, {
method: 'POST',
body: formData,
});
if (!addResponse.ok) {
return { error: `HTTP error ${addResponse.status}: ${addResponse.statusText}` };
}
const addListData = await addResponse.json();
const userListId = addListData.userListId;
if (!userListId) {
return { error: `'userListId' not found in Enrichr's /addList response: ${JSON.stringify(addListData)}` };
}
// 2. Get enrichment results
const enrichUrl = `${ENRICHR_URL}/enrich`;
const geneSetLibrary = "GO_Biological_Process_2025";
const enrichParams = new URLSearchParams({
userListId: userListId.toString(),
backgroundType: geneSetLibrary
});
const enrichResponse = await fetch(`${enrichUrl}?${enrichParams}`);
if (!enrichResponse.ok) {
return { error: `HTTP error ${enrichResponse.status}: ${enrichResponse.statusText}` };
}
const enrichmentResults = await enrichResponse.json();
if (!(geneSetLibrary in enrichmentResults)) {
return { error: `Library '${geneSetLibrary}' not found in results. Available: ${Object.keys(enrichmentResults).join(', ')}` };
}
const allResults = enrichmentResults[geneSetLibrary];
// Filter for significant results using adjusted p-value (adjusted p < 0.05)
// Result format: [rank, term_name, p_value, odds_ratio, combined_score, overlapping_genes, adjusted_p_value, old_p_value, old_adjusted_p_value]
const significantResults = allResults.filter((result) => result[6] < 0.05);
return {
total_terms: allResults.length,
significant_terms: significantResults.length,
results: significantResults
};
}
catch (error) {
return { error: `Error querying Enrichr: ${error instanceof Error ? error.message : String(error)}` };
}
}
/**
* Format enrichment results for readable output
*/
function formatEnrichmentResults(resultsData, maxTerms = CONFIG.maxTermsPerLibrary, format = CONFIG.format) {
if ("error" in resultsData) {
return `Error: ${resultsData.error}`;
}
const { total_terms, significant_terms, results } = resultsData;
if (significant_terms === 0) {
return `No significant GO Biological Process terms found (adjusted p < 0.05) out of ${total_terms} total terms analyzed.`;
}
const actualTermsShown = Math.min(significant_terms, maxTerms);
const outputLines = [
`Found ${significant_terms} significant GO Biological Process terms (adjusted p < 0.05) out of ${total_terms} total terms:`
];
if (significant_terms > maxTerms) {
outputLines.push(`Showing top ${actualTermsShown} terms (use --max-terms to show more):\n`);
}
else {
outputLines.push('');
}
if (format === "compact") {
// Compact format
results.slice(0, maxTerms).forEach((termInfo, i) => {
const [rank, termName, pValue, oddsRatio, combinedScore, overlappingGenes, adjustedPValue] = termInfo;
outputLines.push(`${i + 1}. ${termName} (adj-p: ${adjustedPValue.toExponential(2)}, genes: ${overlappingGenes.length})`);
});
}
else if (format === "minimal") {
// Minimal format: just term name and p-value
results.slice(0, maxTerms).forEach((termInfo, i) => {
const [rank, termName, pValue, oddsRatio, combinedScore, overlappingGenes, adjustedPValue] = termInfo;
outputLines.push(`${i + 1}. ${termName} (p: ${pValue.toExponential(2)})`);
});
}
else {
// Detailed format
results.slice(0, maxTerms).forEach((termInfo, i) => {
const [rank, termName, pValue, oddsRatio, combinedScore, overlappingGenes, adjustedPValue] = termInfo;
outputLines.push(`${i + 1}. ${termName}`);
outputLines.push(` Adjusted P-value: ${adjustedPValue.toExponential(2)}`);
outputLines.push(` Raw P-value: ${pValue.toExponential(2)}`);
outputLines.push(` Odds Ratio: ${oddsRatio.toFixed(2)}`);
outputLines.push(` Combined Score: ${combinedScore.toFixed(2)}`);
outputLines.push(` Overlapping Genes (${overlappingGenes.length}): ${overlappingGenes.join(', ')}`);
outputLines.push("");
});
}
return outputLines.join("\n");
}
/**
* Save enrichment results to TSV file
*/
function saveResultsToTSV(results, filename, description = "Enrichment Analysis") {
try {
const timestamp = new Date().toISOString();
const tsvLines = [];
// Add header
tsvLines.push("# Enrichr Enrichment Analysis Results");
tsvLines.push(`# Generated: ${timestamp}`);
tsvLines.push(`# Description: ${description}`);
tsvLines.push(`# Libraries: ${Object.keys(results).join(', ')}`);
tsvLines.push("#");
// Add column headers
tsvLines.push([
"Library",
"Rank",
"Term_Name",
"Term_ID",
"Adjusted_P_Value",
"Raw_P_Value",
"Odds_Ratio",
"Combined_Score",
"Gene_Count",
"Overlapping_Genes"
].join("\t"));
// Add data rows
for (const [library, result] of Object.entries(results)) {
if ("error" in result) {
tsvLines.push([library, "ERROR", result.error, "", "", "", "", "", "", ""].join("\t"));
continue;
}
result.results.forEach((termInfo) => {
const [rank, termName, pValue, oddsRatio, combinedScore, overlappingGenes, adjustedPValue] = termInfo;
// Extract term ID from term name if it exists (e.g., "Term Name (GO:1234567)")
const termIdMatch = termName.match(/\(([^)]+)\)$/);
const termId = termIdMatch ? termIdMatch[1] : "";
const cleanTermName = termId ? termName.replace(/\s*\([^)]+\)$/, "") : termName;
tsvLines.push([
library,
rank.toString(),
cleanTermName,
termId,
adjustedPValue.toExponential(6),
pValue.toExponential(6),
oddsRatio.toFixed(4),
combinedScore.toFixed(4),
overlappingGenes.length.toString(),
overlappingGenes.join(';')
].join("\t"));
});
}
// Write to file
writeFileSync(filename, tsvLines.join("\n") + "\n");
}
catch (error) {
throw new Error(`ā Error saving TSV file: ${error instanceof Error ? error.message : String(error)}`);
}
}
/**
* Create an MCP server with capabilities for tools only (Enrichr GO enrichment)
*/
const server = new Server({
name: "enrichr-server",
version: CONFIG.version,
}, {
capabilities: {
tools: {},
},
});
/**
* Handler that lists available tools.
* Exposes enrichment analysis tools with multi-library support.
*/
server.setRequestHandler(ListToolsRequestSchema, async () => {
// Dynamically generate the description for the enrichr_analysis tool
const baseDescription = `Perform gene set enrichment analysis using Enrichr with support for multiple gene set libraries.
Use this tool when you need to:
- analyze gene functions
- test enrichment across different databases
- find biological processes/pathways/diseases
- perform functional enrichment
- analyze gene sets
- identify overrepresented terms
- run enrichment analysis
- perform gene ontology analysis
- test for enriched biological terms
- analyze gene list functionality across multiple databases.
Returns only statistically significant terms (adjusted p < 0.05) to reduce context usage.
Supports GO, pathways, disease, tissue, drug, and many other gene set libraries available in Enrichr.`;
const configuredLibrariesDescription = CONFIG.defaultLibraries.map(lib => ` - ${lib}: ${libraryDescriptions[lib] || 'Custom library or no description available.'}`).join('\n');
const dynamicDescription = `${baseDescription}\n\nThis server is configured with the following default libraries:\n${configuredLibrariesDescription}\n\nThe model should select the most relevant library/libraries from the list below based on the user's query.`;
// Generate a list of all available libraries for the parameter description
const allAvailableLibraries = Object.keys(libraryDescriptions).map(lib => `'${lib}'`).join(', ');
return {
tools: [
{
name: "enrichr_analysis",
description: dynamicDescription,
inputSchema: {
type: "object",
properties: {
genes: {
type: "array",
items: {
type: "string"
},
description: "List of gene symbols to analyze for enrichment (e.g., ['TP53', 'BRCA1', 'EGFR'])"
},
libraries: {
type: "array",
items: {
type: "string"
},
description: `List of Enrichr libraries to use for analysis. If not specified, the configured defaults will be used. Available options include: ${allAvailableLibraries}.`,
default: CONFIG.defaultLibraries
},
description: {
type: "string",
description: "Optional description for the gene list",
default: "Gene list for enrichment analysis"
},
maxTerms: {
type: "integer",
description: "Maximum number of terms to show per library. Use higher values (20-50) to capture more biological insights, especially for libraries with many significant terms.",
default: CONFIG.maxTermsPerLibrary,
minimum: 1,
maximum: 100
},
format: {
type: "string",
description: "Output format: detailed, compact, minimal",
default: CONFIG.format,
enum: ["detailed", "compact", "minimal"]
},
outputFile: {
type: "string",
description: "Optional path to save complete results as TSV file. If specified, ALL significant terms will be saved to this file, regardless of maxTerms limit.",
default: CONFIG.outputFile
}
},
required: ["genes"]
}
},
{
name: "go_bp_enrichment",
description: "Perform Gene Ontology (GO) Biological Process enrichment analysis to understand what biological functions and processes are overrepresented in your gene list. This tool helps researchers interpret gene expression data, identify statistically significant biological processes, and uncover functional implications of genes from RNA-seq, microarray, or other high-throughput experiments. Use this when you need to: analyze gene functions, find enriched biological processes, perform functional profiling of gene lists, understand molecular mechanisms, interpret differentially expressed genes (DEGs), discover key biological pathways, annotate gene lists functionally, characterize gene sets involved in specific phenotypes, connect genes to their biological roles, or investigate what your genes do. The tool performs over-representation analysis (ORA) using the Enrichr API and GO Biological Process 2025 database, returning only statistically significant terms (adjusted p-value < 0.05) to provide meaningful biological insights while managing context usage. Perfect for transcriptomics analysis, systems biology studies, drug target identification, biomarker discovery, and understanding disease mechanisms. For multi-library analysis across different databases (KEGG, Reactome, etc.), use enrichr_analysis instead.",
inputSchema: {
type: "object",
properties: {
genes: {
type: "array",
items: {
type: "string"
},
description: "List of gene symbols to analyze for GO BP enrichment (e.g., ['TP53', 'BRCA1', 'EGFR']). Can be from DEG analysis, candidate gene lists, or any gene set of interest."
},
description: {
type: "string",
description: "Optional description for the gene list to help track analyses",
default: "Gene list for GO BP enrichment"
},
outputFile: {
type: "string",
description: "Optional path to save complete results as TSV file. If specified, ALL significant terms will be saved to this file, regardless of maxTerms limit.",
default: CONFIG.outputFile
}
},
required: ["genes"]
}
}
]
};
});
/**
* Handler for enrichment analysis tools
*/
server.setRequestHandler(CallToolRequestSchema, async (request) => {
switch (request.params.name) {
case "enrichr_analysis": {
const genes = request.params.arguments?.genes;
const libraries = request.params.arguments?.libraries || CONFIG.defaultLibraries;
const description = request.params.arguments?.description || "Gene list for enrichment analysis";
const maxTerms = request.params.arguments?.maxTerms || CONFIG.maxTermsPerLibrary;
const format = request.params.arguments?.format || CONFIG.format;
const outputFile = request.params.arguments?.outputFile || CONFIG.outputFile;
if (!genes) {
return {
content: [{
type: "text",
text: "Error: No genes provided. Please provide a list of gene symbols."
}]
};
}
if (!Array.isArray(genes) || !genes.every(gene => typeof gene === "string")) {
return {
content: [{
type: "text",
text: "Error: Genes must be provided as a list of strings (gene symbols)."
}]
};
}
if (!Array.isArray(libraries) || !libraries.every(lib => typeof lib === "string")) {
return {
content: [{
type: "text",
text: "Error: Libraries must be provided as a list of strings (library names)."
}]
};
}
if (genes.length < 2) {
return {
content: [{
type: "text",
text: "Warning: Enrichment analysis typically requires at least 2 genes for meaningful results."
}]
};
}
// Query multiple libraries
const resultsData = await queryEnrichrLibraries(genes, libraries, description);
const formattedResults = formatMultiLibraryResults(resultsData, maxTerms, format);
// Save to TSV if output file is specified (either via CLI or tool parameter)
if (outputFile) {
try {
saveResultsToTSV(resultsData, outputFile, description);
const savedMessage = `\nš¾ Complete results saved to: ${outputFile}`;
return {
content: [{
type: "text",
text: formattedResults + savedMessage
}]
};
}
catch (error) {
const errorMessage = `\nā Error saving TSV file: ${error instanceof Error ? error.message : String(error)}`;
return {
content: [{
type: "text",
text: formattedResults + errorMessage
}]
};
}
}
return {
content: [{
type: "text",
text: formattedResults
}]
};
}
case "go_bp_enrichment": {
const genes = request.params.arguments?.genes;
const description = request.params.arguments?.description || "Gene list for GO BP enrichment";
const outputFile = request.params.arguments?.outputFile || CONFIG.outputFile;
if (!genes) {
return {
content: [{
type: "text",
text: "Error: No genes provided. Please provide a list of gene symbols."
}]
};
}
if (!Array.isArray(genes) || !genes.every(gene => typeof gene === "string")) {
return {
content: [{
type: "text",
text: "Error: Genes must be provided as a list of strings (gene symbols)."
}]
};
}
if (genes.length < 2) {
return {
content: [{
type: "text",
text: "Warning: GO enrichment analysis typically requires at least 2 genes for meaningful results."
}]
};
}
// Query single library (backward compatibility)
const resultsData = await queryEnrichrGoBp(genes, description);
const formattedResults = formatEnrichmentResults(resultsData, CONFIG.maxTermsPerLibrary, CONFIG.format);
// Save to TSV if output file is specified (either via CLI or tool parameter)
if (outputFile) {
try {
// Convert single result to multi-library format for TSV saving
const multiLibraryResults = { "GO_Biological_Process_2025": resultsData };
saveResultsToTSV(multiLibraryResults, outputFile, description);
const savedMessage = `\nš¾ Complete results saved to: ${outputFile}`;
return {
content: [{
type: "text",
text: formattedResults + savedMessage
}]
};
}
catch (error) {
const errorMessage = `\nā Error saving TSV file: ${error instanceof Error ? error.message : String(error)}`;
return {
content: [{
type: "text",
text: formattedResults + errorMessage
}]
};
}
}
return {
content: [{
type: "text",
text: formattedResults
}]
};
}
default:
throw new Error("Unknown tool");
}
});
/**
* Start the server using stdio transport.
* This allows the server to communicate via standard input/output streams.
*/
async function main() {
const transport = new StdioServerTransport();
await server.connect(transport);
}
main().catch((error) => {
console.error("Server error:", error);
process.exit(1);
});