Welcome to LDAK-KVIK

LDAK-KVIK is a powerful and computationally efficient method for mixed-model association analysis in genome-wide association studies (GWAS). It is part of the LDAK software, which is written in C. A preprint of LDAK-KVIK is now online.

🚀 Why Use LDAK-KVIK?

Scalability: Analyzes hundres of thousands of individuals within hours
Statistical power: Improved detection of associated loci using an elastic net model
Handles structure: Robust to relatedness and population stratification
Saddlepoint approximation: Control of type 1 error when testing rare diseases or variants
Gene-based testing: Leverage summary statistics in a gene-based association test

To download LDAK, please visit our downloads page. Example code of LDAK-KVIK can be found on the example code page. An overview of the steps involved in the LDAK-KVIK algorithm is included on the LDAK-KVIK steps page.

Stay updated with the latest developments regarding LDAK software by signing up for the LDAK mailing list.

🔧 Example code

LDAK-KVIK is run in three steps:

Compute polygenic risk scores (PRS) using an elastic net model
Run single-SNP association analysis using PRS as an offset
Conduct gene-based analysis using summary statistics

Example command lines are:

./ldak6.linux --kvik-step1 kvik --bfile data --pheno phenofile --covar covfile --max-threads 4
./ldak6.linux --kvik-step2 kvik --bfile data --pheno phenofile --covar covfile --max-threads 4
./ldak6.linux --kvik-step3 kvik --bfile data --genefile genefile --max-threads 4

The main output files are:

kvik.step2.assoc (results from single-SNP association analysis)
kvik.step3.remls.all (results from gene-based association analysis)

💻 Getting started

LDAK is available for Linux and Mac. On the Downloads page, we list several options for loading or installing the LDAK software. Explore example code and a detailed overview of the KVIK algorithm to get started.

❓ Questions or issues

If you have questions or run into issues, please raise an issue on our GitHub.