Circular Hi-C and genomic data¶

import numpy as np
import matplotlib.pyplot as plt

from circhic import datasets
from circhic import CircHiCFigure

from iced.normalization import ICE_normalization

# Load the data, compute the cumulative raw counts.
data = datasets.load_ccrescentus()
counts_raw = data["counts"]
lengths = data["nbins"]
cumul_raw_counts = counts_raw.sum(axis=0)

# Normalize the data using ICE, and keep the biases
counts, bias = ICE_normalization(counts_raw, output_bias=True)

# compute extreme values

vmax = np.max([counts[i, (i+1) % counts.shape[0]]
               for i in range(counts.shape[0])])
vmin = np.min(counts[counts > 0])

# plotting the data
granularity = 0.5
resolution = 9958

fig = plt.figure(figsize=(6, 6))

inner_radius, outer_radius = 0.005, 0.75
inner_gdis, outer_gdis = 2000000, 600000

chrom_lengths = lengths * resolution
circhicfig = CircHiCFigure(chrom_lengths, figure=fig)

m, ax = circhicfig.plot_hic(counts, granularity=granularity,
                            resolution=resolution, outer_radius=outer_radius,
                            inner_radius=inner_radius, inner_gdis=inner_gdis,
                            outer_gdis=outer_gdis, vmin=vmin*100, vmax=vmax,
                            cmap="bone_r", border_thickness=0.005)

rax = circhicfig.plot_raxis()
rax.set_yticklabels(["200", "0", "60"], fontsize="small")

rax.set_ylabel("Genomic distance (kb)", fontsize="small", color="0.3",
               position=(0, 1.03))
rax.tick_params(colors="0.3")

# Assume you want to plot data from that ranges in a polar plot outside of the
# first one. Then the 0 axis should be at, say, 80% of the axis
bar, _ = circhicfig.plot_bars(
    cumul_raw_counts, inner_radius=0.8, outer_radius=0.9,
    color="C0")

# Now, plot another plot, for the top 10% of the original axes
lines, _ = circhicfig.plot_lines(
    bias, color="C1", inner_radius=0.9, outer_radius=1)

cab = circhicfig.set_colorbar(m, orientation="horizontal")
cab.set_label("Normalized contact counts", fontsize="small")

ticklabels = ["%d kb" % (i * 500) for i in range(6)]
tickpositions = [int(i*500000) for i in range(6)]
ticklabels[0] = "oriC"
ax = circhicfig.set_genomic_ticklabels(
    tickpositions=tickpositions,
    ticklabels=ticklabels,
    outer_radius=1, fontdict={'fontsize': "small"})
ax.tick_params(colors="0.3")

fig.legend((bar, lines[0]), ("Mappability", "Bias"), fontsize="small",
           bbox_to_anchor=(0.8, 0.1, 0.15, 0.15), frameon=False)

print("")