variant_correlation: Process QTL Intervals

Description Usage Arguments Details Value

View source: R/variant_correlation.R

Description

variant_correlation Returns all highly correlated variants in a QTL confidence interval

Usage

1
2
3
variant_correlation(df, quantile_cutoff = 0.9,
  variant_severity = c("MODERATE", "SEVERE"), gene_types = "ALL",
  kin = kinship, condition_trait)

Arguments

df

is a dataframe that is output from the process_mappings function

quantile_cutoff

is a quantile cutoff that determines what variants to keep, default is to keep all variants with correlation coefficients greater than the 90th quantile

variant_severity

what variants to look at from snpeff output

gene_types

what gene types to look at from snpeff output

kin

is a strain by strain relatedness matrix you want to correct you trait with

Details

This function losely identifies unique intervals for QTL in a data set and browses a whole-genome variant set for variants that are highly correlated with the phenotype that gave rise to the QTL. Spearman rank correlation is used. Heterozygotes are ignored in the analysis. Correlations are only calculated for variants that are present in at least 5 acquired for at least 80

Value

Outputs a list. Each list contains two data frames, the first contains mapping information (e.g. log10p, confidence interval start and stop), phenotype information, and gene ids. The second element of the list contains more detailed gene information


AndersenLab/cegwas documentation built on March 6, 2020, 1:10 a.m.