pyGenomeViz

Overview

pyGenomeViz is a genome visualization python package for comparative genomics implemented based on matplotlib. This package is developed for the purpose of easily and beautifully plotting genomic features and sequence similarity comparison links between multiple genomes. It supports genome visualization of Genbank format file and can be saved figure in various formats (JPG/PNG/SVG/PDF). User can use pyGenomeViz for interactive genome visualization figure plotting on jupyter notebook, or automatic genome visualization figure plotting in genome analysis scripts/pipelines.

For more information, please see full documentation here.

Fig.1 Four Erwinia phage genome comparison result

Fig.2 Six Enterobacteria phage genome comparison result

Installation

Python 3.7 or later is required for installation.

Install PyPI package:

pip install pygenomeviz

Install bioconda package:

conda install -c conda-forge -c bioconda pygenomeviz

Examples

Jupyter notebooks containing code examples below is available here.

Basic Example

Single Track

from pygenomeviz import GenomeViz

name, genome_size = "Tutorial 01", 5000
cds_list = ((100, 900, -1), (1100, 1300, 1), (1350, 1500, 1), (1520, 1700, 1), (1900, 2200, -1), (2500, 2700, 1), (2700, 2800, -1), (2850, 3000, -1), (3100, 3500, 1), (3600, 3800, -1), (3900, 4200, -1), (4300, 4700, -1), (4800, 4850, 1))

gv = GenomeViz()
track = gv.add_feature_track(name, genome_size)
for idx, cds in enumerate(cds_list, 1):
    start, end, strand = cds
    track.add_feature(start, end, strand, label=f"CDS{idx:02d}")

gv.savefig("example01.png")

Multiple Tracks & Links

from pygenomeviz import GenomeViz

genome_list = (
    {"name": "genome 01", "size": 1000, "cds_list": ((150, 300, 1), (500, 700, -1), (750, 950, 1))},
    {"name": "genome 02", "size": 1300, "cds_list": ((50, 200, 1), (350, 450, 1), (700, 900, -1), (950, 1150, -1))},
    {"name": "genome 03", "size": 1200, "cds_list": ((150, 300, 1), (350, 450, -1), (500, 700, -1), (701, 900, -1))},
)

gv = GenomeViz(tick_style="axis")
for genome in genome_list:
    name, size, cds_list = genome["name"], genome["size"], genome["cds_list"]
    track = gv.add_feature_track(name, size)
    for idx, cds in enumerate(cds_list, 1):
        start, end, strand = cds
        track.add_feature(start, end, strand, label=f"gene{idx:02d}", linewidth=1, labelrotation=0, labelvpos="top", labelhpos="center", labelha="center")

# Add links between "genome 01" and "genome 02"
gv.add_link(("genome 01", 150, 300), ("genome 02", 50, 200))
gv.add_link(("genome 01", 700, 500), ("genome 02", 900, 700))
gv.add_link(("genome 01", 750, 950), ("genome 02", 1150, 950))
# Add links between "genome 02" and "genome 03"
gv.add_link(("genome 02", 50, 200), ("genome 03", 150, 300), normal_color="skyblue", inverted_color="lime", curve=True)
gv.add_link(("genome 02", 350, 450), ("genome 03", 450, 350), normal_color="skyblue", inverted_color="lime", curve=True)
gv.add_link(("genome 02", 900, 700), ("genome 03", 700, 500), normal_color="skyblue", inverted_color="lime", curve=True)
gv.add_link(("genome 03", 900, 701), ("genome 02", 1150, 950), normal_color="skyblue", inverted_color="lime", curve=True)

gv.savefig("example02.png")

Exon Features

from pygenomeviz import GenomeViz

exon_regions1 = [(1, 210), (301, 480), (591, 800), (851, 1000), (1031, 1300)]
exon_regions2 = [(1501, 1710), (2001, 2480), (2591, 2800)]
exon_regions3 = [(3001, 3300), (3401, 3690), (3801, 4100), (4201, 4610)]

gv = GenomeViz()
track = gv.add_feature_track(name=f"Exon Features", size=5000)
track.add_exon_feature(exon_regions1, strand=1, plotstyle="box", label="box", labelrotation=0, labelha="center")
track.add_exon_feature(exon_regions2, strand=-1, plotstyle="arrow", label="arrow", labelrotation=0, labelha="center", facecolor="darkgrey", intron_patch_kws={"ec": "red"})

exon_labels = [f"exon{i+1}" for i in range(len(exon_regions3))]
track.add_exon_feature(exon_regions3, strand=1, plotstyle="bigarrow", label="bigarrow", facecolor="lime", linewidth=1, exon_labels=exon_labels, labelrotation=0, labelha="center", exon_label_kws={"y": 0, "va": "center", "color": "blue"})

gv.savefig("example03.png")

Practical Example

Single Track from Genbank file

from pygenomeviz import Genbank, GenomeViz, load_dataset

gbk_files, _ = load_dataset("enterobacteria_phage")
gbk = Genbank(gbk_files[0])

gv = GenomeViz()
track = gv.add_feature_track(gbk.name, gbk.genome_length)
track.add_genbank_features(gbk)

gv.savefig("example04.png")

Multiple Tracks & Links from Genbank files

from pygenomeviz import Genbank, GenomeViz, load_dataset

gv = GenomeViz(
    fig_track_height=0.7,
    feature_track_ratio=0.2,
    tick_track_ratio=0.7,
    tick_style="bar",
    align_type="center",
)

gbk_files, links = load_dataset("escherichia_phage")
for gbk_file in gbk_files:
    gbk = Genbank(gbk_file)
    track = gv.add_feature_track(gbk.name, gbk.genome_length)
    track.add_genbank_features(gbk, facecolor="limegreen", linewidth=0.5, arrow_shaft_ratio=1.0)

for link in links:
    link_data1 = (link.ref_name, link.ref_start, link.ref_end)
    link_data2 = (link.query_name, link.query_start, link.query_end)
    gv.add_link(link_data1, link_data2, v=link.identity, curve=True)

gv.savefig("example05.png")

Customization Tips

Since pyGenomeViz is implemented based on matplotlib, users can easily customize the figure in the manner of matplotlib. Here are some tips for figure customization.

Customization Tips 01

• Add GC Content & GC skew subtrack
• Add annotation label & fillbox
• Add colorbar for links identity

from pygenomeviz import Genbank, GenomeViz, load_dataset

gv = GenomeViz(
    fig_width=12,
    fig_track_height=0.7,
    feature_track_ratio=0.5,
    tick_track_ratio=0.3,
    tick_style="axis",
    tick_labelsize=10,
)

gbk_files, links = load_dataset("erwinia_phage")
gbk_list = [Genbank(gbk_file) for gbk_file in gbk_files]
for gbk in gbk_list:
    track = gv.add_feature_track(gbk.name, gbk.genome_length, labelsize=15)
    track.add_genbank_features(gbk, plotstyle="arrow")

min_identity = int(min(link.identity for link in links))
for link in links:
    link_data1 = (link.ref_name, link.ref_start, link.ref_end)
    link_data2 = (link.query_name, link.query_start, link.query_end)
    gv.add_link(link_data1, link_data2, v=link.identity, vmin=min_identity)

# Add subtracks to top track for plotting 'GC content' & 'GC skew'
gv.add_feature_subtrack(gv.top_track.name, "gc_content", ratio=0.7)
gv.add_feature_subtrack(gv.top_track.name, "gc_skew", ratio=0.7)

fig = gv.plotfig()

# Add label annotation to top track
top_track = gv.top_track  # or, gv.get_track("MT939486") or gv.get_tracks()[0]
label, start, end = "Inverted", 310000 + top_track.offset, 358000 + top_track.offset
center = int((start + end) / 2)
top_track.ax.hlines(1.5, start, end, colors="red", linewidth=1, linestyles="dashed", clip_on=False)
top_track.ax.text(center, 2.0, label, fontsize=12, color="red", ha="center", va="bottom")

# Add fillbox to top track
x, y = (start, start, end, end), (1, -1, -1, 1)
top_track.ax.fill(x, y, fc="lime", linewidth=0, alpha=0.1, zorder=-10)

# Plot GC content for top track
gc_content_ax = gv.top_track.subtracks[0].ax  # or, gv.get_track("gc_content").ax
pos_list, gc_content_list = gbk_list[0].calc_gc_content()
gc_content_ax.set_ylim(bottom=0, top=max(gc_content_list))
pos_list += gv.top_track.offset  # Offset is required if align_type is not 'left'
gc_content_ax.fill_between(pos_list, gc_content_list, alpha=0.2, color="blue")
gc_content_ax.text(gv.top_track.offset, max(gc_content_list) / 2, "GC(%) ", ha="right", va="center", color="blue")

# Plot GC skew for top track
gc_skew_ax = gv.top_track.subtracks[1].ax  # or, gv.get_track("gc_skew").ax
pos_list, gc_skew_list = gbk_list[0].calc_gc_skew()
gc_skew_abs_max = max(abs(gc_skew_list))
gc_skew_ax.set_ylim(bottom=-gc_skew_abs_max, top=gc_skew_abs_max)
pos_list += gv.top_track.offset  # Offset is required if align_type is not 'left'
gc_skew_ax.fill_between(pos_list, gc_skew_list, alpha=0.2, color="red")
gc_skew_ax.text(gv.top_track.offset, 0, "GC skew ", ha="right", va="center", color="red")

# Set coloarbar for link
gv.set_colorbar(fig, vmin=min_identity)

fig.savefig("example06.png", bbox_inches="tight")

Customization Tips 02

• Add legends
• Add colorbar for links identity

from matplotlib.lines import Line2D
from matplotlib.patches import Patch

from pygenomeviz import Genbank, GenomeViz, load_dataset

gv = GenomeViz(
    fig_width=10,
    fig_track_height=0.7,
    feature_track_ratio=0.5,
    tick_track_ratio=0.5,
    align_type="center",
    tick_style="bar",
    tick_labelsize=10,
)

gbk_files, links = load_dataset("enterobacteria_phage")
for idx, gbk_file in enumerate(gbk_files):
    gbk = Genbank(gbk_file)
    track = gv.add_feature_track(gbk.name, gbk.genome_length, labelsize=10)
    track.add_genbank_features(
        gbk,
        label_type="product" if idx == 0 else None,  # Labeling only top track
        label_handle_func=lambda s: "" if s.startswith("hypothetical") else s,  # Ignore 'hypothetical ~~~' label
        labelsize=8,
        labelvpos="top",
        facecolor="skyblue",
        linewidth=0.5,
    )

normal_color, inverted_color, alpha = "chocolate", "limegreen", 0.5
min_identity = int(min(link.identity for link in links))
for link in links:
    link_data1 = (link.ref_name, link.ref_start, link.ref_end)
    link_data2 = (link.query_name, link.query_start, link.query_end)
    gv.add_link(link_data1, link_data2, normal_color, inverted_color, alpha, v=link.identity, vmin=min_identity, curve=True)

fig = gv.plotfig()

# Add Legends (Maybe there is a better way)
handles = [
    Line2D([], [], marker=">", color="skyblue", label="CDS", ms=10, ls="none"),
    Patch(color=normal_color, label="Normal Link"),
    Patch(color=inverted_color, label="Inverted Link"),
]
fig.legend(handles=handles, frameon=True, bbox_to_anchor=(1, 0.8), loc="upper left", ncol=1, handlelength=1, handleheight=1)

# Set colorbar for link
gv.set_colorbar(fig, bar_colors=[normal_color, inverted_color], alpha=alpha, vmin=min_identity, bar_height=0.15, bar_label="Identity", bar_labelsize=10)

fig.savefig("example07.png", bbox_inches="tight")