martinjzhang/ldspec

Gene-level directional effect regression

LDSPEC

LDSPEC (LD SNP-pair effect correlation regression) is a method for estimating the correlation of causal disease effect sizes for pairs of nearby SNPs, depending on their functional annotations.

Reference

Zhang, et al. "Pervasive correlations between causal disease effects of proximal SNPs vary with functional annotations and implicate stabilizing selection", preprint, 2023.

Versions

• v0.0.4: beta version accompanying the initial submission.

Code and data for the paper

See experiments for more details. Data are at figshare.

• Download baseline-SP annotations (165 single-SNP annotations and 136 SNP-pair annotations, 14,820,648 imputed SNPs).
• Download corresponding LD scores and directional LD scores (14,820,648 imputed SNPs)
• Download GWAS summary statistics for 70 UK Biobank diseases/traits (14,820,648 imputed SNPs).

Using LDSPEC

Installation

git clone https://github.com/martinjzhang/LDSPEC.git 
cd LDSPEC
pip install -e .

Usage

Please check out this self-contained example for steps of LDSPEC and inputs and outputs in each step.

LDSPEC consists of many steps and may be tricky to run. We are working hard to make it more user-friendly. Let us know if you have any questions or suggestions for the software.

Contact

• Martin Zhang ([email protected]).