SPAGRM is a scalable, accurate, and universal analysis framework to control for sample relatedness in large-scale genome-wide association studies (GWAS). In the paper A scalable, accurate, and universal analysis framework to control for sample relatedness in large-scale genome-wide association studies and its application to 79 longitudinal traits in UK Biobank (to be updated), we applied SPAGRM to analyze 79 longitudinal traits extracted from UK Biobank primary care data. As a universal analysis framework, we also evaluated SPAGRM's performance in quantitative and binary trait analysis.
SPAGRM is now implemented in the GRAB package. Please click here to download.
Detailed documentation about how to use SPAGRM is available at SPAGRM documentation.
Summary statistics of 79 longitudinal traits extracted from UK Biobank primary care data is available at here.
This repository contains:
- old version of SPAGRM;
- figures of GWAS results of 79 longitudinal traits;
- materials for reproducing the experiments.
Scripts to reproduce the experiments performed for the SPAGRM manuscript:
A scalable, accurate, and universal analysis framework to control for sample relatedness in large-scale genome-wide association studies and its application to 79 longitudinal traits in UK Biobank (to be updated)
- extract longitudinal traits from UK Biobank primary care data
In real_data\1.extract_pheno:
# phecode-2023-05-10XH.R defines the Read v2 and CTV3 code terms for 79 longitudinal traits.
# real_data_extra-2023-05-01XH.R preprocesses the gp_clinical table.
- preprocess longitudinal traits
In real_data\2.preprocess:
# XXX_preprocess.R extracts longitudinal traits from gp_clinical table and preprocesses each longitudinal traits.
- fit the null model via WiSER package
In real_data\3.null_fitter:
# XXX_null_fitter.jl fits the null model and obtains model residuals for each longitudinal trait.
- GWAS analysis
- SPAGRM analysis
In real_data\4.1.SPAGRM:
# XXX_SPAGRM.R conducts the SPAGRM analysis with two parts: 1). pre-calculate of the joint distribution of genotypes and 2). conduct genome-wide association studies for each SNP.
- TrajGWAS analysis
In real_data\4.2.TrajGWAS:
# XXX_TrajGWAS.jl conducts the TrajGWAS analysis.
- PRS adjustment (optional)
In real_data\5.PRS_adjustment (optional):
# 1.select_index_SNPs_from_results_of_SPAGRM.R selects independent SNPs that pass the given p value threshold based on the GWAS results of SPAGRM.
# 2.estimate_effect_sizes_XXX.jl estimate the effect sizes of index SNPs via WiSER package.
# 3.construct_LOCO-PRS.R constructs the Leave One Chromosome Out Polygenic Risk Scores (LOCO-PRS).
# 4.adjust_PRS_in_null_model_fitting_XXX.jl adds the PRS into the null model fitting.
# 5.second_round_of_GWAS_via_SPAGRM.R conducts the second round of GWAS via SPAGRM.
In simulation:
# 1.simupheno contains R script to simulate longitudinal data.
# 2.typeIerror contains scripts to conduct 1e9 type I error simulations.
# 3.power contains scripts to conduct 1e3 power simulations.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent