feat: QTL Analysis Skill - Open Source Alternative to QTLmax

[borealBytes]

Summary

Create comprehensive QTL analysis Agent Skill as an open-source alternative to commercial QTLmax software. This skill provides unified access to industry-standard QTL/GWAS tools (tensorQTL, GEMMA, PLINK, R/qtl2) through a single CLI interface.

Motivation

QTLmax is a commercial Linux desktop application for QTL analysis. This skill brings equivalent functionality using open-source tools that are:

• Free and open-source
• Well-documented and peer-reviewed
• GPU-accelerated (tensorQTL)
• Industry-standard (GEMMA, PLINK, R/qtl2)

Scope

Core Components

• SKILL.md with complete documentation
• Unified CLI (qtl_cli.py) with 6 subcommands
• Preflight checker (check_system.py) for dependencies
• One-shot installer (install_deps.sh)

Examples (4 complete workflows)

• GWAS-LMM: GEMMA GWAS with Manhattan + QQ plots
• eQTL-cis: tensorQTL eQTL with LocusZoom
• Classical-QTL: R/qtl2 LOD curves for F2 crosses
• Population Structure: PCA + admixture + kinship heatmap

Reference Documentation

• API reference (tensorQTL, GEMMA, PLINK, R/qtl2)
• Data formats (VCF, PLINK, phenotypes)
• Cross types guide (F2, BC, RIL, DO, MPP)
• Tool comparison matrix

Example Assets

• Input data files (CSV format)
• Output visualizations (PNG format)
• Example READMEs with Input→Process→Output structure

Project Board

Todo

• Add animated workflow diagrams (Mermaid/GIF)
• Create integration tests with pytest
• Add GitHub Actions CI for example validation

In Progress

• Code review and final verification

Done

• Create branch feat/qtl-analysis-skill
• Write SKILL.md (351 lines)
• Build unified CLI entry point
• Create 4 comprehensive examples
• Generate example outputs and visualizations
• Write complete reference documentation
• Add example READMEs with acceptance criteria

Tools Coverage

Tool	Purpose	Status
tensorQTL	cis-eQTL mapping	✅
GEMMA	GWAS with LMM	✅
PLINK	QC, PCA, conversion	✅
R/qtl2	Classical QTL, MPP	✅

Files Added

scientific-skills/qtl-analysis/
├── SKILL.md
├── scripts/
│   ├── check_system.py
│   ├── qtl_cli.py
│   └── install_deps.sh
├── examples/
│   ├── gwas-lmm/
│   │   ├── run_gwas.py
│   │   ├── README.md
│   │   └── output/
│   ├── eqtl-cis/
│   │   ├── run_eqtl.py
│   │   ├── README.md
│   │   └── output/
│   ├── classical-qtl/
│   │   ├── run_lodscan.py
│   │   ├── README.md
│   │   └── output/
│   └── population-structure/
│       ├── run_structure.py
│       ├── README.md
│       └── output/
└── references/
    ├── api_reference.md
    ├── data_formats.md
    ├── cross_types.md
    └── tool_comparison.md

Branch

feat/qtl-analysis-skill from borealBytes/claude-scientific-skills

Related Resources

• QTLmax reference: https://open.qtlmax.com/guide/
• R/qtl2: https://kbroman.org/qtl2/
• tensorQTL: https://github.com/broadinstitute/tensorqtl
• GEMMA: https://github.com/genetics-statistics/GEMMA

[rekram1-node]

closing due to issue spam