Oncoprint (Python)

🤖 AI Skill — Download the Bizard unified skill for your AI assistant ⬇️ Download Skill ZIP

🔌 MCP Server — Let AI agents (Claude Code, Cursor, …) query Bizard live: claude mcp add bizard -- npx @openbiox/bizard-mcp — or the free hosted instance (no install, faster from mainland China): claude mcp add --transport http bizard http://biotree.top:38123/bizard-mcp/mcp 📖 MCP Docs

📖 Cite Bizard — if Bizard helps your research, please cite: Li, K., Zheng, H., Huang, K., et al. (2026). Bizard: A Community-Driven Platform for Accelerating and Enhancing Biomedical Data Visualization. iMetaMed, e70038. 🔗 DOI / BibTeX

An oncoprint tiles one gene (row) against one tumour sample (column) per altered cell, colour-coded by alteration type, with per-gene frequency and per-sample mutation burden in flanking bars — the signature figure of tumour genomics papers. matplotlib builds it from one coloured-rectangle pass plus two bar axes; the hand-rolled version keeps every sorting rule explicit.

Example

Setup

• System Requirements: Cross-platform (Linux/MacOS/Windows)
• Programming Language: Python
• Dependencies: matplotlib, seaborn, pandas, numpy

import matplotlib.pyplot as plt
import seaborn as sns
import pandas as pd
import numpy as np

Data Preparation

We simulate a 12-gene × 80-tumour alteration table (Missense, Truncating, Amplification) with realistic frequency skew — TP53 near-ubiquitous, drivers rare but recurrent — plus per-sample mutation counts.

np.random.seed(261)

genes = ['TP53', 'KRAS', 'STK11', 'KEAP1', 'EGFR', 'BRAF',
         'PIK3CA', 'ALK', 'ERBB2', 'MET', 'RB1', 'SMARCA4']
freq = np.array([0.46, 0.30, 0.17, 0.15, 0.13, 0.08,
                 0.07, 0.05, 0.05, 0.04, 0.09, 0.06])
n_tumours = 80

alt_records = []
for gene, f in zip(genes, freq):
    for tumour in np.random.choice(n_tumours, size=int(f * n_tumours),
                                   replace=False):
        alt_records.append(
            (gene, tumour, np.random.choice(
                ['Missense', 'Truncating', 'Amplification'],
                p=[0.55, 0.2, 0.25])))
alts = pd.DataFrame(alt_records, columns=['Gene', 'Tumour', 'Type'])

type_colors = {'Missense': '#66c2a5', 'Truncating': '#e63946',
               'Amplification': '#8da0cb'}
alt_order = ['Truncating', 'Missense', 'Amplification']  # draw order

Visualization

Sorted Oncoprint with Flanking Bars

Genes are sorted by alteration frequency (top row = most altered), samples by total burden (left column = most loaded), and both flanking bars quantify the same two axes.

gene_order = alts.Gene.value_counts().index.tolist()
burden = alts.groupby('Tumour').size().reindex(range(n_tumours), fill_value=0)
sample_order = burden.sort_values(ascending=False).index.tolist()

fig, (ax_top, ax_mat, ax_right) = plt.subplots(
    3, 1, figsize=(8.6, 5.4), sharex=False,
    gridspec_kw={'height_ratios': [1, 4, 0.6], 'hspace': 0.04})

counts = alts.Gene.value_counts().reindex(gene_order)
ax_top.bar(range(len(gene_order)), counts.values, color='#457b9d',
           edgecolor='0.25', linewidth=0.4)
ax_top.set_ylabel('Altered')
ax_top.spines[['top', 'right']].set_visible(False)

for y, gene in enumerate(gene_order):
    sub = alts[alts.Gene == gene]
    for _, row in sub.iterrows():
        x = sample_order.index(row.Tumour)
        ax_mat.add_patch(plt.Rectangle(
            (x - 0.5, y - 0.42), 1.0, 0.84,
            color=type_colors[row.Type], linewidth=0))
ax_mat.set_yticks(range(len(gene_order)), gene_order)
ax_mat.invert_yaxis()
ax_mat.set_xlim(-0.5, n_tumours - 0.5)
ax_mat.spines[['top', 'right', 'left', 'bottom']].set_visible(False)
ax_mat.tick_params(length=0)
ax_mat.set_ylabel('Gene')

ax_right.barh(range(len(sample_order)),
              burden.loc[sample_order].values, color='#457b9d',
              edgecolor='0.25', linewidth=0.3, height=0.8)
ax_right.invert_yaxis()
ax_right.set_xlabel('Mutations per tumour')
ax_right.spines[['top', 'right', 'left']].set_visible(False)

handles = [plt.Rectangle((0, 0), 1, 1, color=c) for c in type_colors.values()]
ax_top.legend(handles, type_colors, frameon=False, ncol=3,
              loc='lower left', bbox_to_anchor=(0.0, 1.02), fontsize=8)
plt.show()

Figure 1: Alteration landscape across 12 drivers in 80 tumours

Highlighting One Alteration Type

Drawing non-Missense events with a thin outline keeps the dominant alteration class readable while still showing where truncating events cluster.

fig, (ax_top, ax_mat) = plt.subplots(
    2, 1, figsize=(8.6, 4.6), sharex=False,
    gridspec_kw={'height_ratios': [1, 4], 'hspace': 0.05})

ax_top.bar(range(len(gene_order)), counts.values, color='#457b9d',
           edgecolor='0.25', linewidth=0.4)
ax_top.set_ylabel('Altered')
ax_top.spines[['top', 'right']].set_visible(False)

for y, gene in enumerate(gene_order):
    sub = alts[alts.Gene == gene]
    for _, row in sub.iterrows():
        x = sample_order.index(row.Tumour)
        is_trunc = row.Type == 'Truncating'
        ax_mat.add_patch(plt.Rectangle(
            (x - 0.5, y - 0.42), 1.0, 0.84,
            facecolor=type_colors[row.Type] if is_trunc else '#d9dee6',
            edgecolor=type_colors[row.Type] if is_trunc else 'none',
            linewidth=1.2 if is_trunc else 0))
ax_mat.set_yticks(range(len(gene_order)), gene_order)
ax_mat.invert_yaxis()
ax_mat.set_xlim(-0.5, n_tumours - 0.5)
ax_mat.spines[['top', 'right', 'left', 'bottom']].set_visible(False)
ax_mat.tick_params(length=0)
ax_mat.set_ylabel('Gene')
plt.show()

Figure 2: Oncoprint emphasising truncating alterations

References

1. Gao, J., et al. (2013). Integrating genomics data into the cBioPortal. Science Signaling, 6(268), pl4.
2. Hunter, J. D. (2007). Matplotlib: A 2D graphics environment. Computing in Science & Engineering, 9(3), 90-95.