Nature期刊图复现 | Python 绘制UpSet 交集矩阵点阵图与柱状图组合【常用于替代传统的韦恩图(VD)】

import osimport pandas as pdimport numpy as npimport matplotlib.pyplot as pltos.makedirs('图表', exist_ok=True)if not os.path.exists('data.xlsx'):    raise FileNotFoundError("未在同目录下找到 data.xlsx 文件，请确保数据文件已生成。")df_culture = pd.read_excel('data.xlsx', sheet_name='culture_genes')df_inter_up = pd.read_excel('data.xlsx', sheet_name='intersection_up')df_inter_down = pd.read_excel('data.xlsx', sheet_name='intersection_down')samples = df_culture['Culture'].tolist()# 20种专业科研配色方案COLOR_PALETTES = {    1: {'up': '#B22222', 'down': '#104E8B', 'gray': '#C0C0C0'},  # 经典红蓝：Nature/Science高频使用，对比鲜明}# 当前选择：方案 1SELECTED_PALETTE_INDEX = 1palette = COLOR_PALETTES[SELECTED_PALETTE_INDEX]color_up_main = palette['up']color_down_main = palette['down']color_gray = palette['gray']color_dot_bg = '#EAEAEA'color_blue_dot = '#00BFFF'fig = plt.figure(figsize=(12, 8), dpi=300)def draw_upset_sub(fig, rect_top, rect_matrix, rect_side, df_inter, culture_vals, is_up=True):    main_color = color_up_main if is_up else color_down_main    ax_top = fig.add_axes(rect_top)    ax_top.bar(samples, culture_vals, color=main_color, width=0.6, edgecolor='black', linewidth=0.8)    ax_top.set_ylabel('Genes per culture', fontsize=11)    ax_top.set_title('Upregulated' if is_up else 'Downregulated', fontsize=13, loc='left', pad=10, weight='bold')    ax_top.spines['top'].set_visible(False)    ax_top.spines['right'].set_visible(False)    ax_top.spines['bottom'].set_visible(False)    ax_top.tick_params(axis='x', which='both', bottom=False, top=False, labelbottom=False)    ax_top.tick_params(axis='y', labelsize=10)    ax_top.yaxis.set_major_formatter(plt.FuncFormatter(lambda x, loc: "{:,}".format(int(x))))    ax_top.set_ylim(0, 2800 if is_up else 3800)    ax_matrix = fig.add_axes(rect_matrix)    n_rows = len(df_inter)    n_cols = len(samples)    for i in range(n_rows):        ax_matrix.axhspan(i - 0.4, i + 0.4, color=color_dot_bg, alpha=0.5, zorder=0)        row_status = [df_inter.iloc[i][col] for col in samples]        active_idx = [j for j, val in enumerate(row_status) if val == 1]        if len(active_idx) > 1:            ax_matrix.plot([active_idx[0], active_idx[-1]], [i, i], color='black', linewidth=1.5, zorder=1)        for j in range(n_cols):            if row_status[j] == 1:                dot_color = color_blue_dot if i == 0 else 'black'                ax_matrix.scatter(j, i, color=dot_color, s=70, zorder=2)            else:                ax_matrix.scatter(j, i, color=color_gray, s=70, zorder=2)    ax_matrix.set_xlim(-0.5, n_cols - 0.5)    ax_matrix.set_ylim(-0.5, n_rows - 0.5)    ax_matrix.set_xticks(range(n_cols))    ax_matrix.set_xticklabels(samples, rotation=90, fontsize=11)    ax_matrix.set_yticks([])    for spine in ax_matrix.spines.values():        spine.set_visible(False)    ax_matrix.tick_params(axis='both', which='both', length=0)    ax_matrix.invert_yaxis()    ax_side = fig.add_axes(rect_side)    sizes = df_inter['Intersection_Size'].tolist()    colors_side = [main_color] + [color_gray] * (n_rows - 1)    ax_side.barh(range(n_rows), sizes, color=colors_side, height=0.6, edgecolor='black', linewidth=0.8)    ax_side.set_xlabel('No. of intersections', fontsize=11)    ax_side.set_ylim(-0.5, n_rows - 0.5)    ax_side.spines['top'].set_visible(False)    ax_side.spines['right'].set_visible(False)    ax_side.spines['left'].set_visible(False)    ax_side.set_yticks([])    ax_side.tick_params(axis='y', which='both', left=False, right=False, labelleft=False)    ax_side.tick_params(axis='x', labelsize=10)    ax_side.invert_yaxis()    target_y = 1.8    if is_up:        ax_side.set_xlim(0, 28)        ax_side.annotate('', xy=(sizes[0] + 4.45, target_y - 0.55), xytext=(sizes[0] + 0.5, 0.2),                         arrowprops=dict(arrowstyle="->", color="black", lw=0.8), zorder=5)        text_up = "$CLDN4$\n$TM4SF1$\n$ZFAS1$"        ax_side.text(sizes[0] + 4.5, target_y, text_up,                     bbox=dict(boxstyle="square,pad=0.5", fc="white", ec="gray", ls="--", lw=0.8),                     fontsize=10, va='center', style='italic', zorder=4)    else:        ax_side.set_xlim(0, 52)        ax_side.annotate('', xy=(sizes[0] + 6.45, target_y - 0.55), xytext=(sizes[0] + 0.5, 0.2),                         arrowprops=dict(arrowstyle="->", color="black", lw=0.8), zorder=5)        text_down = "$TFF2$\n$LYZ$\n$SPTBN1$"        ax_side.text(sizes[0] + 6.5, target_y, text_down,                     bbox=dict(boxstyle="square,pad=0.5", fc="white", ec="gray", ls="--", lw=0.8),                     fontsize=10, va='center', style='italic', zorder=4)rect_up_top = [0.06, 0.61, 0.20, 0.25]rect_up_matrix = [0.06, 0.15, 0.20, 0.45]rect_up_side = [0.27, 0.15, 0.15, 0.45]rect_down_top = [0.47, 0.61, 0.20, 0.25]rect_down_matrix = [0.47, 0.15, 0.20, 0.45]rect_down_side = [0.68, 0.15, 0.15, 0.45]draw_upset_sub(fig, rect_up_top, rect_up_matrix, rect_up_side, df_inter_up, df_culture['Upregulated_Genes'], is_up=True)draw_upset_sub(fig, rect_down_top, rect_down_matrix, rect_down_side, df_inter_down, df_culture['Downregulated_Genes'],               is_up=False)fig.text(0.01, 0.94, 'e', fontsize=18, weight='bold')output_file = '图表/intersection_upset_plot.png'plt.savefig(output_file, bbox_inches='tight', dpi=300)plt.close()print(f"绘图完成。图片已保存至: {output_file}")