Manhattan plot — manhattan

Create a genome-wide Manhattan plot from GWAS summary statistics. Returns a ggplot2 object for further customization.

Usage

manhattan_plot(
  data,
  chr = NULL,
  bp = NULL,
  p = NULL,
  snp = NULL,
  colors = c("#1A5276", "#76D7C4"),
  point_size = 0.8,
  alpha = 1,
  genome_wide = 5e-08,
  suggestive = 1e-05,
  threshold_colors = c("#E74C3C", "#3498DB"),
  highlight_snps = NULL,
  highlight_color = "#E74C3C",
  highlight_size = 2,
  label_snps = NULL,
  label_top_n = NULL,
  label_column = "SNP",
  downsample = TRUE,
  downsample_n = 2e+05,
  chromosomes = NULL,
  chr_labels = NULL,
  y_metric = "p",
  y_limit = NULL,
  y_truncate = NULL,
  title = NULL
)

Arguments

data: A gwas_data object or data.frame with GWAS results.
chr, bp, p, snp: Column name overrides (auto-detected if NULL).
colors: Two-color vector for alternating chromosomes.
point_size: Point size.
alpha: Point transparency.
genome_wide: Genome-wide significance threshold (p-value).
suggestive: Suggestive significance threshold.
threshold_colors: Colors for significance lines.
highlight_snps: Character vector of SNP IDs to highlight.
highlight_color: Color for highlighted SNPs.
highlight_size: Size for highlighted points.
label_snps: Character vector of SNP IDs to label.
label_top_n: Label the top N most significant SNPs.
label_column: Column to use for label text.
downsample: Enable smart downsampling for large datasets.
downsample_n: Target number of points after downsampling.
chromosomes: Subset of chromosomes to plot (integer vector).
chr_labels: Custom chromosome labels. Options: NULL (all labels), "odd" (only odd-numbered chromosomes labeled), or a character vector of labels (same length as displayed chromosomes).
y_metric: What to plot on the y-axis. "p" (default) shows -log10(p). "beta_min" shows the lower confidence bound of the absolute effect size: |beta| - 2*SE. Variants whose confidence interval overlaps zero are excluded. This highlights variants with large, robust effect sizes rather than just small p-values.
y_limit: Upper y-axis limit for -log10(p).
y_truncate: Break the y-axis for datasets with extreme peaks. A single value shows 0 to that value at full scale and compresses everything above it into a band above the break (nothing is dropped), e.g. y_truncate = 15. A vector of two values c(break_from, resume_at) instead cuts out the middle region: y_truncate = c(15, 50) shows 0-15 at full scale, drops 15-50, then shows 50+ above the break. Units are -log10(p).
title: Plot title.

Value

A ggplot object.

Examples

data(example_gwas, package = "ggwas")

# Basic Manhattan plot
manhattan_plot(example_gwas)


# Label top hits with a different palette
manhattan_plot(example_gwas, label_top_n = 5, colors = gwas_palette("vibrant"))


# Nature journal style
manhattan_plot(example_gwas, label_top_n = 3) + theme_nature()


# Lancet palette with Science theme
manhattan_plot(example_gwas, colors = gwas_palette("lancet")) + theme_science()


# Subset chromosomes
manhattan_plot(example_gwas, chromosomes = seq_len(10))


# NEJM palette
manhattan_plot(example_gwas, colors = gwas_palette("nejm"), label_top_n = 3)


# Broken y-axis: cut 10-50, show 0-10 and 50+
manhattan_plot(example_gwas, y_truncate = c(10, 50))


# Single value: auto-detect resume point
manhattan_plot(example_gwas, y_truncate = 10)


# Custom significance threshold (Bonferroni for 500k SNPs)
manhattan_plot(example_gwas, genome_wide = 0.05 / 500000)


# No threshold lines
manhattan_plot(example_gwas, genome_wide = NULL, suggestive = NULL)


# Label only odd chromosomes (less crowded x-axis)
manhattan_plot(example_gwas, chr_labels = "odd")


# Effect-size confidence bound (|beta| - 2*SE)
manhattan_plot(example_gwas, y_metric = "beta_min")