README.md

In adamwaring/ClusterBurden: Rare variant association utilising burden and positional information.

ClusterBurden

To install in R:

if(!"devtools" %in% installed.packages()[,"Package"]){\ install.packages("devtools")\ }\ library(devtools)\ install_github("adamwaring/ClusterBurden")

Due to some large variant and coverage files, download may be slow. if interest in specific functions, feel free to copy directly from the R/ folder in this gitHub repo.

Rare variant association using both positional (variant amino acid residue clustering) and burden signals - with control for uneven coverage.

Relevance of method:

This method is designed for case-control studies where the condition under investigation is caused by rare missense variants in protein-coding regions. This covers many Mendelian/rare diseases. When a simple burden test does not yield significant associations, consideration of clustering information can increase power to detect low penetrance undiscovered disease-genes.

In this package

This package allows exome-wide scans searching for both burden and clustering signals as well as their combined significance. GnomAD controls (exome and genomes v2) are available to automatically generate a control set and their coverage files to adjust for uneven coverage. Once downloaded (some moderately large files present) exome-wide scans are quick. See the associated vignettes for more details about the methods, how to conduct exome-wide scans using automatic controls and formatting user data.

Functions overview

Statistical methods (all with coverage control)

• BIN-test: detect positional differences of missense variants between cases and controls
• burden_test: detect burden differences of missense variants between cases and controls
• combine_ps: combine p-values from independent statistical tests
• BIN_test_WES: whole-exome scans for clustering signals
• burden_test_WES: whole-exome scans for burden signals
• ClusterBurden_WES: whole-exome scans for burden, clustering and combined signals

Visualisation

• plot_distribs: plot variant and coverage distributions for a single gene
• plot_signif_distribs: plot variant and coverage distributions for top significant genes in WES analysis
• manhhattan: manhattan plot for p-values from WES analysis
• plot_features: plot variant positions alongside sequence features from uniprot
• plot_residuals: plot standardized residuals from a chi-squared two-sample test

Formatting inputs

• collect_gnomad_controls: generate automatic controls based on GnomAD exomes or genomes v2
• remove_uncovered_residues: exclude residue positions from analysis which have poor coverage in cases or controls
• get_residue_position: extract variant residue position from HGVSp protein consequence format
• annotate_with_gnomad_freqs: annotate a dataset with frequencies from GnomAD exomes and genomes for filtering
• format_coverage: map a chr-pos coverage file into a gene-protein_index coverage file (canonical transcripts only)

adamwaring/ClusterBurden documentation built on July 29, 2020, 9:50 p.m.