mtDNA-CN-ExWAS

Rare variant association testing to mitochondrial copy number in UK Biobank

Author: Vamsee Pillalamarri

This repo contains code (pipeline, scripts, and analysis notebooks) relating to Pillalamarri et al. 2023 HGG Advances "Whole-exome sequencing in 415,422 individuals identifies rare variants associated with mitochondrial DNA copy number" (PMID 36311265, doi: 10.1016/j.xhgg.2022.100147).

This repository is a snapshot of a private repository (ArkingLab/mito_rare-variant) with more comprehensive access to a variety of resources for the project along with results from the single- and rare-variant burden testing. Please reach out to [email protected] or [email protected] for raw BOLT-LMM results from single- and GENESIS aggregate rare-variant burden testing results.

Study Design

R Package Requirements

Install BIOCONDUCTOR Install Manager:

if (!require("BiocManager", quietly = TRUE))
    install.packages("BiocManager")

Aggregate Association Testing

• GENESIS
• SeqArray

Install Bioconductor packages:

BiocManager::install("SeqArray")
BiocManager::install("GENESIS")

Scripts

1. Aggregate Association Testing using GENESIS: Example call:

   Rscript src/scripts/aggregate_association_testing.R 1 --AF.max 0.01 --test SMMAT --qv nonsyn.impact_mod_high --ncores 4
   

2. Leave one out testing: Example call:

   Rscript src/scripts/leave_one_out.R --gene GENE_NAME --chr CHROMOSOME --af_max 0.01 --output results.rds --plot --plot_type SMMAT --save_dir ./plots --cores 4
   

3. Power analysis:

   • src/scripts/power-analysis.R
   • src/scripts/power-analysis.py (Rewritten in Python)

4. Gene-level conditional variant breakdown functions:

   • conditional_variants.breakdown.R

5. Single-variant testing using BOLT-LMM:

• src/scripts/pipeline_shell_scripts/run.bolt_lmm_adjusted_forGWASsnps.sh

Run as:
# On UKB RAP CLI:
#  <log into UKB RAP using `dx login`>
#       qrsh -now y -pe local 16 -l h_vmem=5G,mem_free=5G,h_fsize=10G
# OPTIONAL:
#       cd /dcs01/arking/arkinglab/software/src/BOLT-LMM/
#       module load singularity/3.2
#       module load python/
#       singularity shell -B /tmp,/dcl01,/dcs01 singularity-bolt-lmm.simg

Analysis Notebooks

1. Table 1A: src/analysis_notebooks/Table_1A.Rmd
2. Table 1B src/analysis_notebooks/Table_1B.Rmd
3. PheWAS Analyses src/analysis_notebooks/phewas-analyses.Rmd
4. PheWAS Additional Analyses src/analysis_notebooks/phewas-additional-analyses.Rmd
5. Model Genetic Architecture src/analysis_notebooks/model-genetic-architecture.Rmd

Data Locations

• pVCFs: /dcl01/arking/data/static/UKBiobank/WES/pVCF
• plink: /dcl01/arking/data/static/UKBiobank/WES/plink

NOTE: pVCFs combine multiallelic sites into one record, while plink splits them; the plink .bim file reflects this difference

Gene Lists

Gene lists can be found in the "resources" folder in this repo, or on JHPCE at /dcl01/arking/data/static/UKBiobank/gene_lists.

• gwas_genes.txt: Genes identified from GWAS for mtDNA-CN
• depletion_genes.txt: 15 genes known to be causal for mtDNA depletion syndromes

Mitocarta Gene Pathways

Mitocarta gene pathways: /dcs01/arking/arkinglab/resources/mitoCarta/Human.MitoPathways3.0.gmx