Meta-MR: Mendelian Randomization Meta-Analysis Platform

Generate harmonized JSON data and perform comprehensive MR analysis

Zhang WJ. 2027. Meta-MR: The meta-methodology and the web platform for Mendelian Randomization analysis. Network Biology, 17(2): 121-467


Step A0: Choose Your Data Workflow

Option 1: Direct JSON Input

Already have harmonized JSON data? Paste it directly and you're done. Perfect for pre-processed datasets.

Option 2: File Upload & Harmonization

Upload exposure/outcome files (CSV, TXT, JSON, GZ), harmonize with correlation matrix from AI or public databases, get JSON output.

Option 3: AI and Multi-Source Solutions

Use AI to fetch GWAS data by trait names OR fetch real GWAS data from multiple databases (IEU OpenGWAS, GWAS Catalog, FinnGen, GWAS Atlas, eQTLGen).

Direct JSON Input

Paste your existing harmonized JSON data below:

Expected Format: JSON object with keys: snps (array), betaX, betaY, betaXse, betaYse (arrays), correlation (2D array), exposure, outcome (strings). Alternatively, you can directly edit the JSON data in the box.

File Upload & Harmonization

Upload your exposure and outcome files, then harmonize with correlation matrix:

Step 1: Upload Files

Click or drag to upload

CSV format: SNP,ea,nea,beta,se,eaf
JSON format: {"rs9939609": {"ea":"A","nea":"T","beta":0.15,"se":0.02}}

Click or drag to upload

CSV format: SNP,beta,se,ea,nea,eaf
JSON format: {"rs9939609": {"beta":-0.08,"se":0.03,"ea":"A","nea":"T"}}

Step 2: Harmonization & Correlation Matrix

Correlation Matrix Source: Choose between public genetic databases or AI-fetched correlation matrices.
DeepSeek: Get API key (Free tier available)
Gemini: Get API key

AI and Multi-Source Solutions

Choose between AI fetched GWAS data or fetching from multiple databases:

Step 1: Analysis Configuration

For univariate analysis, enter one trait. For multivariate analysis, enter multiple traits separated by commas.

Step 2: AI Configuration for Data Fetching

DeepSeek: Get API key
Gemini: Get API key
GPT: Get API key

AI is fetching GWAS data... This may take 30-60 seconds.

Step 2: Data Sources

Select one or more databases to search for GWAS summary statistics.

Step 3: API Authentication (Optional)

Some databases may require authentication tokens for access or higher rate limits.
SNPs with p-values below this threshold will be considered genome-wide significant.