pyryotype

Table of Contents

• pyryotype
  • Acknowledgements
  • Installation
  • Example usage
    • You want vertical? We got vertical
    • Coverage plotting
    • PAF plotting
  • CNV from bam
  • Bandage friendly CSV
  • License
  • Cytoband data

Acknowledgements

Huge shout out to Ryan Dale (@daler) for the original work on the Ideograms - https://gist.github.com/daler/c98fc410282d7570efc3 And @alexomics for the ground work on the coverage plotting and ideograms.

Installation

    pip install ideogram

Example usage

from pyryotype import GENOME, plot_ideogram
from matplotlib import pyplot as plt
fig, axes = plt.subplots(
    ncols=1,
    nrows=22,
    figsize=(11, 11),
    facecolor="white",
)
genome = GENOME.CHM13
for ax, contig_name in zip(axes, range(1, 23)):
    chromosome = f"chr{contig_name}"
    plot_ideogram(ax, target=chromosome, genome=genome)
fig.savefig("ideogram.png", dpi=300)

Will output:

You want vertical? We got vertical

With Regions highlighted on Chromosome 1 - Optional! See the following testing code snippet for an example def test_23_vertical_chm13_regions()

Will output

Coverage plotting

Coverage plotting is designed to be used with the output of Mosdepth. The following example uses the output of mosdepth to plot the coverage of chromosome 1. The region representing the first 100Mb of the chromosome is highlighted in black.

from pathlib import Path
from matplotlib import pyplot as plt
from matplotlib.ticker import EngFormatter
import pandas as pd

from pyryotype.coverage import plot_coverage


test_cov_file = Path(__file__).parent / "static" / "chr1_cov.regions.bed.gz"

fig, ax = plt.subplots(
    ncols=1,
    nrows=1,
    figsize=(11, 2),
)
df = pd.read_csv(test_cov_file, sep="\t", names=["chromosome", "start", "end", "value"])
ax = plot_coverage(ax, df, "chr1", regions=[(0, 100000000)], ylabel="Coverage", color="black")
ax.set_xlabel("Genomic Position (bp)")

# ax.yaxis.set_visible(False)
ax.set_yticks([])
ax.set_yticklabels([])
for side in ("right", "top", "left"):
    ax.spines[side].set_visible(False)
formatter = EngFormatter(unit='b', places=1)
# Set formatter for the x-axis
ax.xaxis.set_major_formatter(formatter)

fig.savefig("example_outputs/test_coverage.png", dpi=300, bbox_inches="tight")

Will output something like:

PAF plotting

Designed primarily to plot the alignment of assemblies to a reference genome. Must provide a valid PAF file.

from matplotlib import pyplot as plt
from readpaf import parse_paf
from pathlib import Path
from pyryotype.paf_plotting import PAFProtocol, PlotMode, plot_paf_alignments
test_paf = Path(__file__).parent / "static" / "test.paf"
fig, ax = plt.subplots(
    ncols=1,
    nrows=1,
    figsize=(11, 1),
)

ax = plot_paf_alignments(
    ax,
    parse_paf(test_paf.open()),
    target="chr1",
    mapq_filter=0,
    strict=PlotMode.STRICT,
    contig_colours=PlotMode.UNIQUE_COLOURS,
)
ax.set_xlabel("")

# ax.yaxis.set_visible(False)
ax.set_yticks([])
ax.set_yticklabels([])
for side in ("right", "top", "left"):
    ax.spines[side].set_visible(False)
fig.savefig("tests/test_paf_plotting.png", dpi=300, bbox_inches="tight")

Will output the following image: The colours assigned to each alignment can be changed to either be based on the Alignment Strand, or unique for each record. See the PlotMode docstring for the options that can be applied.

Supplementary mappings on the same chromosome/strand for a given sequence can be collapsed into a single line by setting strict=PlotMode.STRICT. This can be seen in the above mappings, where the large block is comprised of 3 separate alignments, from the same read. IF strict=PlotMode.CHILL then each alignment will be plotted separately, even if these alignments are from the same read. This looks like:

There are now more options! Chevrons representing the alignment strand can be added by setting chevron=PlotMode.CHEVRONS in the plot_paf_alignments function. Chevrons will only be drawn if the rendered alignment block is wide enough. The fontsize of the chevron can be set with kwarg chevron_fontsize. An example image be seen in the following example:

We can now expand multiple alignments that overlap onto separate tracks (a la IGV.) This can be down by setting expand=PlotMode.EXPAND and strict=PlotMode.STRICT. This will expand overlapping alignments onto separate "tracks" on an axis. The number of tracks for a given axis can be set by the kwarg argument max_tracks. If there are more overlapping alignments that there are tracks, the track with the smalles overlap will be chosen. This can be seen in the following example (which also has chevrons enabled):

CNV from bam

A separate repo calculates CNV from BAM files by binning alignment mapping starts and returning CNV by a dynamically calculated bin width. See here for more detail - https://github.com/Adoni5/cnv_from_bam

Bandage friendly CSV

For visualising Genomes that provide GFA, we can take a PAF file of the genome aligned to a suitable reference,a nd output a CSV for labelling the sequence nodes in bandage. See https://github.com/Adoni5/pyryotype/blob/ec2239cd0fac6ff57e3129cc52ac860d729c94e5/src/pyryotype/paf_tools.py#L114 for more.

[!NOTE] Note: For alignments with many supplementary alignments, the alignment is collapsed to the alignment that covers the largest area of a single contig on the reference.

License

pyryotype is distributed under the terms of the MIT license.

Cytoband data

• HG38 - Nushell, will have to be adapted for bash curl -L "https://hgdownload.cse.ucsc.edu/goldenpath/hg38/database/cytoBand.txt.gz" | gzip -d - | rg -Ne "^chr\\d+\t" | save cytoBand_HG38.tsv
• CHM13 - bash yay curl -L http://t2t.gi.ucsc.edu/chm13/hub/t2t-chm13-v2.0/download/chm13v2.0_cytobands_allchrs.bed.gz | gzip -d - > cytobands_chm13.bed