lamm-mit/skills/gtex-database

GTEx Database Skill Summary

Purpose

This skill enables querying the Genotype-Tissue Expression (GTEx) portal to investigate tissue-specific gene expression, expression quantitative trait loci (eQTLs), and splicing QTLs—critical for connecting GWAS variants to regulatory mechanisms.

Primary Use Cases

The documentation identifies several key applications:

• Linking non-coding GWAS variants to regulated genes via eQTL analysis
• Comparing expression patterns across 54 human tissues
• Testing whether GWAS and eQTL signals share causal variants
• Identifying variants affecting RNA splicing ratios

Core Technical Components

API Foundation: The GTEx REST API v2 (base: https://gtexportal.org/api/v2/) provides JSON responses without authentication and supports pagination.

Key Endpoints:

• Gene expression lookup returns median TPM values per tissue
• eQTL queries retrieve significant variant-gene associations with p-values and effect sizes
• Variant-based searches identify all genes regulated by a specific SNP
• Tissue lists provide available samples with metadata

Data Format: The skill uses GENCODE identifiers for genes and employs a standardized variant ID format: chr{chrom}_{pos}_{ref}_{alt}_b38.

Practical Implementation

The documentation provides Python examples demonstrating pagination handling, DataFrame filtering, and multi-tissue aggregation. It emphasizes using GRCh38 coordinates and notes that FDR < 0.05 defines GTEx significance thresholds.

License

The skill is distributed under CC-BY-4.0, authored by Kuan-lin Huang.