Pixi Binary Installation Guide (🥈)

Overview

Pixi Binary installation provides the fastest installation method for Linux and MacOS users with pre-built packages that require no compilation. This method is ideal for users who want a quick setup without dealing with compilers or build dependencies.

Why Pixi Binary?

✅ Advantages:

Fastest installation (no compilation required)
No compiler or build tools needed
Avoids compilation errors completely
Pre-built optimized binaries
Consistent environment management
Works on most modern Linux systems

⚠️ Requirements:

Linux x86_64 or MacOS arm64
R 4.4+ (managed by pixi)
GLIBC 2.28+ (CentOS 7+, Ubuntu 18.04+, most modern systems)
Pixi package manager

Quick Installation

1. Install pixi (if needed)

curl -fsSL https://pixi.sh/install.sh | bash
export PATH="$HOME/.pixi/bin:$PATH" # Restart shell or reload environment

2. Download SAIGE-QTL repository

git clone https://github.com/weizhou0/SAIGEQTL.git
cd SAIGEQTL

3. Install from pre-built binary

For Linux users:

# Auto-detect latest binary version
BINARY_FILE=$(ls binaries/SAIGEQTL_*_linux-x86_64.tgz | head -n1)
echo "Installing: $BINARY_FILE"

# Install binary package
CONDA_OVERRIDE_GLIBC=2.28 pixi run R -e "
install.packages('${BINARY_FILE}', repos = NULL, type = 'source')
library(SAIGEQTL)
cat('✓ SAIGEQTL', as.character(packageVersion('SAIGEQTL')), 'installed successfully\n')
"

For MacOS users:

# Auto-detect latest binary version
BINARY_FILE=$(ls binaries/SAIGEQTL_*_macos.tgz | head -n1)
echo "Installing: $BINARY_FILE"

# Install binary package
CONDA_OVERRIDE_GLIBC=2.28 pixi run R -e "
install.packages('${BINARY_FILE}', repos = NULL, type = 'source')
library(SAIGEQTL)
cat('✓ SAIGEQTL', as.character(packageVersion('SAIGEQTL')), 'installed successfully\n')
"

What Gets Installed

Pre-built Binary Package

SAIGEQTL Binary: Pre-compiled Linux or MacOS package (.tgz format)
Location: binaries/SAIGEQTL_*_linux-x86_64.tgz and binaries/SAIGEQTL_*_macos.tgz in the downloaded SAIGEQTL repository
No compilation: Ready-to-install binary eliminates build time and dependencies

Managed R Environment

R Version: 4.4+ with optimized BLAS/LAPACK libraries
Core Dependencies: Automatically managed by pixi environment
Cross-platform Compatibility: Consistent behavior across Linux distributions

Built-in Dependencies

The binary package includes all required dependencies:

Statistical genetics: MetaSKAT, SKAT, SPAtest
Data manipulation: data.table, dplyr, dbplyr
Parallel computing: RcppParallel, furrr
Numerical computing: RcppArmadillo, RcppEigen, RcppNumerical
System libraries: boost-cpp, zlib, zstd, savvy

Repository Access

Repository: https://github.com/weizhou0/qtl
Version: Latest stable release (automatically detected)

Verification

After installation, verify the setup:

# Check if pixi environment is working
CONDA_OVERRIDE_GLIBC=2.28 pixi run R --version

# Test SAIGEQTL installation
CONDA_OVERRIDE_GLIBC=2.28 pixi run R -e 'library(SAIGEQTL); packageVersion("SAIGEQTL")'
# Expected output: [1] '0.3.5' (or latest version)

# Test core functionality
CONDA_OVERRIDE_GLIBC=2.28 pixi run R -e 'library(SAIGEQTL); exists("fitNULLGLMM_multiV")'
# Expected output: [1] TRUE

# Test help information when calling script for step 1 
CONDA_OVERRIDE_GLIBC=2.28 pixi run Rscript ./extdata/step1_fitNULLGLMM_qtl.R --help

✅ Next Steps

Installation Complete!

You’ve successfully installed SAIGEQTL with Pixi Binary method. The package is now ready to use with the fastest possible setup.

Your Command Prefix

All SAIGEQTL commands will use this prefix:

CONDA_OVERRIDE_GLIBC=2.28 pixi run Rscript [script_name] [options]

💡 Tip: Set an alias to save typing:

# Add to your ~/.bashrc or ~/.bash_profile or ~/.zshrc
alias saige-pixi='CONDA_OVERRIDE_GLIBC=2.28 pixi run'

# Then you can run:
saige-pixi Rscript your_path_to_SAIGEQTL_folder/SAIGEQTL/extdata/step1_fitNULLGLMM_qtl.R --help

Quick Start Tutorial

Ready to run your first analysis? Follow this tutorial:

📖 cis-eQTL Analysis Tutorial - Learn how to map local genetic effects on gene expression

What you’ll learn:

How to prepare your data files
Running Step 1: Fit the null model
Running Step 2: Test genetic variants
Running Step 3: Calculate gene-level p-values

Additional Resources

Parameters & Options - Complete reference for all command-line parameters
Running Scripts Guide - Detailed information on running SAIGE-QTL scripts
FAQ - Common questions and troubleshooting

Need Help?

If you encounter issues:

Check the FAQ for common problems
Verify your installation using the verification commands above
Review the cis-eQTL tutorial for step-by-step guidance