nldsc — (non)-additive LD Score Regression

This project is an extension of the LD Score Regression (ldsc), a widely-used method for estimating the heritability of complex traits using summary statistics from genome-wide association studies (GWAS). This extension allows to take into account non-additive effects (but we consider only dominance for now), which have been shown to play a significant role in the genetic architecture of many complex traits.

Implementation Details

• The ldscore module has been written from scratch, but we took into account key ideas from the original ldsc package.
• The h2 module heavily builds on the original implementation and has mostly been refactored.
• The console interface is completely different from ldsc, with only the names of some flags being saved.

Precomputed LD Scores

You can find an .tar.gz archive with additive and non-additive LD Scores pre-calculated on UK Biobank data ($N=315599$) in the "Releases" tab.

Getting Started

Prerequisites

• Python>=3.8
• CMake>=3.20, Ninja
• Armadillo, gfotran, BLAS/OpenBLAS, LAPACK

Ubuntu 20.04, 22.04:

sudo apt update
sudo apt install -y ninja-build gfortran libblas-dev liblapack-dev libarmadillo-dev

Installation

Install manually

Please ensure that you have installed all required packages. Then, clone the repository:

git clone --recursive https://github.com/bayarpark/nldsc.git
cd nldsc
pip install -r requirements.txt
make build

Install via Conda (Linux and MacOS only)

%under development%

Usage

LD Score Estimation

python nldsc ld --bfile <prefix/to/[.bed|.bim|.fam]> --ld-wind-cm 1 -maf 0.0001 --std-thr 1e-5 --out <prefix/to/result>

Heritability Estimation

python nldsc h2 --sumstats <path/to/ldsc-like.sumstats> --ref-ld <path/to/ld> --w-ld <path/to/ld> 

LICENSE

This project is licensed under the GNU GPL-3.0. See License File

Authors

Bayar Park (Novosibirsk State University)

This package (especially the h2 module) is largely based on the sources of the ldsc package written by Brendan Bulik-Sullivan (Broad Institute of MIT and Harvard) and Hilary Finucane (MIT Department of Mathematics)