ezGenomeTracks

Minimal, ggplot2-based genome browser tracks for R. Provides geoms (geom_*) and easy wrappers (ez_*) to plot common genomics data (coverage, genes, features/peaks, interactions/arcs, Hi-C, Manhattan) and compose them vertically with a shared genomic x-axis via vstack_plot().

Note

ezGenomeTracks is my experiment with Vibe Coding. I mostly use Github Copilot Business. I have also tried Cursor, TRAE and Windsurf.

Note

ezGenomeTracks is in early development, many functionalities may be missing or incomplete. Feedback and contributions are welcome!

Installation

# install.packages("devtools")
devtools::install_github("Zepeng-Mu/ezGenomeTracks")

(Optionally install Bioconductor dependencies you need, e.g. GenomicRanges, rtracklayer.)

Core Concepts

1. Region specification: always parse a string like "chr1:100000-101000".
2. Wrappers ez_*() accept data frames, file paths, or lists and apply the proper geom + theme.
3. All tracks must share identical x-limits (same region) for composition.
4. Standard columns follow GRanges style: seqnames, start, end, strand; plus type-specific columns (score, gene annotations, interaction endpoints).

Quick Start

library(ezGenomeTracks)
region <- "chr1:100000-101000"

Coverage / Signal Track

A numeric score over intervals.

cov_df <- tibble::tibble(
  seqnames = "chr1",
  start = seq(100000, 100990, by = 10),
  end = start + 9,
  score = rnorm(length(start), 10, 2)
)
track_cov <- ez_coverage(cov_df, region = region)

Gene / Transcript Track

Exon segments with strand.

genes_df <- tibble::tibble(
  seqnames = "chr1",
  start = c(100050, 100200),
  end   = c(100150, 100350),
  strand = factor(c("plus", "plus"), levels = c("minus", "plus", "Unknown")),
  type = "exon",
  gene_id = "GENE1",
  transcript_id = "TX1"
)
track_gene <- ez_gene(genes_df, region = region)

Feature / Peak Track

Intervals with optional scores (height / color).

features_df <- tibble::tibble(
  seqnames = "chr1",
  start = c(100120, 100300),
  end   = c(100180, 100360),
  score = c(12, 8)
)
track_feat <- ez_feature(features_df, region = region)

Interaction / Arc Track

Pairs of genomic intervals (same or different chromosomes) drawn as arcs.

interactions_df <- tibble::tibble(
  chr1 = "chr1", start1 = 100120, end1 = 100160,
  chr2 = "chr1", start2 = 100300, end2 = 100340,
  score = 5
)
track_arc <- ez_arc(interactions_df, region = region)

Manhattan Track

Association results: point positions + score (e.g. -log10(p)).

manhattan_df <- tibble::tibble(
  seqnames = rep("chr1", 50),
  start = seq(100000, 100000 + 49*200, by = 200),
  end = start + 1,
  score = rexp(50, rate = 0.3)
)
track_manh <- ez_manhattan(manhattan_df, region = region)

Hi-C / Contact Matrix Track

Binned pairwise contacts rendered as a triangular heatmap.

# Minimal synthetic example (x/y bins with a value)
hic_df <- tidyr::expand_grid(
  bin1 = seq(100000, 100000 + 4*200, by = 200),
  bin2 = seq(100000, 100000 + 4*200, by = 200)
) |> dplyr::filter(bin2 >= bin1) |> dplyr::mutate(
  score = runif(dplyr::n(), 0, 1),
  seqnames = "chr1"
) |> dplyr::rename(start1 = bin1, start2 = bin2)
track_hic <- ez_hic(hic_df, region = region)

Composing Multiple Tracks

Stack tracks vertically with synchronized genomic x-axis.

combined <- vstack_plot(
  list(
    coverage = track_cov,
    genes = track_gene,
    peaks = track_feat,
    arcs = track_arc
  ),
  region = region,
  heights = c(2, 1, 1, 1)
)
combined

Input Flexibility

Each ez_* can take: - Data frame (as above) - File path (e.g. BED, bedGraph, GTF) – internally imported with region filtering - Named list of objects (merged into a single track)

Customization

Control theme variants (ez_theme(), track-specific themes), y-axis visibility (none, simple, full), colors (standard ggplot2 scales or custom palettes), and relative track heights in vstack_plot().

Data Conventions Summary

• Coordinates: seqnames, start, end, optional strand (factor with levels minus, plus, Unknown).
• Signal/coverage: score.
• Genes: gene_id, gene_name (optional), transcript_id, type (gene / exon).
• Features/peaks: intervals + optional score.
• Interactions/arcs: chr1, start1, end1, chr2, start2, end2, optional score.
• Hi-C: binned pairwise positions (package converts to triangle), numeric score.

Tips / Pitfalls

• Always use one region string for all tracks to align x-limits.
• Ensure strand factor levels if using gene track.
• Large files: region filtering reduces import time.
• Keep column names exact; wrappers rely on them.

Citation

If this package aids your work, please cite it appropriately (e.g. mention the GitHub repo in Methods section).

License

MIT license.