Make sideways manhattan plot for building locus zoom. Receives output from a single gwas model.

Usage

plot_panvar_manhattan(
  panvar.table.list = NULL,
  gwas.res = NULL,
  ld.list = NULL,
  pvals.in.log = TRUE,
  plot.r2.thresh = 0.2,
  unplotted.alpha = 0.4,
  window,
  sig.line = 3,
  orient = c("H", "V"),
  point.shape.variable = NULL,
  point.shape.scale = NULL,
  point.fill.variable.c = NULL,
  point.fill.scale.c = NULL,
  point.fill.variable.d = NULL,
  point.fill.scale.d = NULL,
  plot.text.size = 11,
  plot.legend.size = 1.2,
  snp.highlight.df = NULL,
  snp.highlight.point.size = 4,
  snp.highlight.shape.var = NULL,
  snp.highlight.shape.scale = NULL,
  snp.highlight.color.var = NULL,
  snp.highlight.color.scale = NULL
)

Arguments

panvar.table.list: list, output from make_panvar_tables. Provide either this list or both gwas.res and ld.list.
gwas.res: data.frame of all gwas results, should contain columns (CHR, POS, PVAL), corresponding to (chromosome, physical position, and pvalue).
ld.list: list, output of get_ld_in_window
pvals.in.log: boolean, are pvalues in input data.frames in -log10(p)?
plot.r2.thresh: minimum LD with qtl snps to plot snps colored by LD
unplotted.alpha: numeric, number from 0 to 1 to indicate alpha values of snps below the plot.r2.thresh. To not plot these snps set value to 0.
window: numeric, kilobases on either side of top QTL snp to plot
sig.line: numeric, -log10(p) value to draw line on plot
orient: character, will rotate plot 90 degrees. vertical (V) or horizontal (H) refers to how the "buildings" of the plot are plotted. "V" places pvalue on y-axis, "H" places pvalues on x-axis.
point.shape.variable: character, column in gwas.res that contains qualitative annotations to be mapped to point shapes. For example impact grades from snpeff. See format_snpeff_annotations. Only accepts up to 5 classes. "IMPACT" and "IMPACT_PLUS" are special cases that will have a pre-assigned scale used if supplied here.
point.shape.scale: ggplot scale, an object with a stored call to ggplot2::scale_shape_manual. More often an output of the function make_consistent_scale.
point.fill.variable.c: character, column in gwas.res that contains quantitative annotations to be plotted as a continuous variable mapped to point fill. For example, variant effect scores. Only provide either continuous or discrete quantitative annotations.
point.fill.scale.c: character or scale object, either a character indicating the option parameter passed to ggplot2::scale_fill_viridis_b that alters the color scale used. Or a previous call to a ggplot2 continuous fill scale for example ggplot2::scale_fill_stepsn.
point.fill.variable.d: character, column in gwas.res that contains annotations to be plotted as a discrete variable mapped to point fill. For example, Year or Trial if combining multiple gwas results. Only provide either continuous or discrete quantitative annotations.
point.fill.scale.d: character or scale object, either a character indicating the option parameter passed to ggplot2::scale_fill_viridis_d that alters the color scale used. Or a previous call to a ggplot2 discrete fill scale for example ggplot2::scale_fill_discrete.
plot.text.size: numeric, size of text in plot.
plot.legend.size: numeric, size of legend. Passed to ggplot2::theme, legend.key.size.
snp.highlight.df: data.frame, table of specific snps that will be highlighted. By default the point size will be larger and they will be plotted red. Data.frame should contain columns (CHR, POS, PVAL). PVAL column format should reflect pvals.in.log.
snp.highlight.point.size: numeric, size of points plotted from snp.highlight.df.
snp.highlight.shape.var: character, column in snp.highlight.df to be mapped to shape of points. Only one of point.shape.variable or snp.highlight.shape.var should be provided.
snp.highlight.shape.scale: ggplot scale, an object with a stored call to ggplot2::scale_shape_manual. More often an output of the function make_consistent_scale. Works best with shapes 15-20 whose color aesthetics map to the entire shape instead of just the outline.
snp.highlight.color.var: character, column in snp.highlight.df to be mapped to color of highlighted points.
snp.highlight.color.scale: ggplot scale, an object with a stored call to ggplot2::scale_shape_manual. More often an output of the function make_consistent_scale.

Value

GGplot of manhattan plot with points colored by R2. Optionally color or change shape of points by other variables. Plots SNP density as 'rug plot' along bottom of manhattan.

Examples

# organize options
tag.snp <- "Chr_05-6857045"
gwas.df <- read.csv(system.file(
    "extdata",
    "PanvarExample_GLM_GWASresults.csv",
    package = "panvaR"))
annotation.table <- read.csv(system.file(
    "extdata",
    "Setaria_shattering_annotation.csv",
    package = "panvaR"))
plink.path <- bigsnpr::download_plink2()
temp.dir <- file.path(tempdir(), "panvar_ex")
dir.create(temp.dir, showWarnings = FALSE)
geno.bed.filename <- "Setaria_shattering_example_pruned.bed"
geno.bed.directory <- system.file("extdata", package="panvaR")

# make input tables
tables <- make_panvar_tables(
  gwas.res = gwas.df,
  tag.snp = tag.snp,
  annotation.table = annotation.table,
  plink.path = plink.path,
  pvals.in.log = F,
  geno.bed.filename = geno.bed.filename,
  geno.bed.directory = geno.bed.directory,
  window = 25,
  temp.dir = temp.dir,
  compute.scores = FALSE,
  snp.to.gene.buffer = 0)
#> Calculating LD
#> Generating snp to gene correspondence
  
# make plot
plot_panvar_manhattan(
  panvar.table.list = tables,
  pvals.in.log = FALSE,
  window = 25,
  sig.line = 6)


# flip it vertical if you want
plot_panvar_manhattan(
  panvar.table.list = tables,
  pvals.in.log = FALSE,
  window = 25,
  sig.line = 6,
  orient = "V")


# clean up
unlink(temp.dir, recursive = TRUE)