GWASTools (version 1.18.0)

qqPlot: QQ plot for genome wide assocation studies

Description

Generates a Quantile-Quantile plot for -log10 p-values from genome wide association tests.

Usage

qqPlot(pval, truncate = FALSE, ylim = NULL, thinThreshold = NULL, ci=TRUE, ...)

Arguments

pval

Vector of p-values

truncate

Either a logical value indicating whether the y-axis should be truncted to the same range as the x-axis, or a numeric value indicating where to truncate the y-axis. See details.

ylim

Limits for the y axis. Ignored if truncate=TRUE or truncate is numeric.

thinThreshold

if not NULL, -log10(pval) threshold for thinning points.

ci

logical indicator for whether to add confidence intervals to plots?

...

Other parameters to be passed directly to plot.

Details

The function generates a Quantile-Quantile plot of p-values on a -log10 scale, with the option of truncating the y-axis to the range of the x-axis (0, -log10(1/length(pval)). If the y-axis is truncated, then points off the top of the plot are denoted by triangles at the upper edge. The 95% confidence interval is shaded in gray.

If truncate is set to a numeric value, then ylim is set to c(0, truncate) only if the value of truncate is bigger than the maximum -log10(pval). (Use the ylim argument if alternatve behavior is desired.)

If requested with thinThreshold, points with p-values < -log10(thinThreshold) are thinned before plotting. All points with -log10(pval) >= thinThreshold plus 10,000 points with -log10(pval) < thinThreshold (randomly selected in uniformly-spaced bins of -log10(pval)) are displayed.

Examples

pvals <- seq(0, 1, 0.001)
qqPlot(pvals)
qqPlot(pvals, thinThreshold=2)
qqPlot(pvals, truncate=TRUE)
qqPlot(pvals, truncate=10)