获取特定节点形状、边颜色和边权重时出错

Question

我想绘制一组基因的有向图。假设有些基因是癌基因，有些是驱动基因。此外，基因与基因的相互作用使用特定的形状和颜色进行加权和表示。我使用以下代码来绘制有向图：

    import networkx as nx
    import matplotlib.pyplot as plt
    import numpy as np
    
    # Your weighted adjacency matrix (replace this with your actual matrix)
    adjacency_matrix = np.array([
        [0, 0.2, 0, 0, 0.4],
        [0.0, 0, 0, 0, 0.1],
        [0.1, 0, 0, 0.1, 0],
        [0, 0, 0.3, 0, 0],
        [0.0, 0.0, 0, 0, 0]
    ])
    
    # Your Katz centrality scores (replace this with your actual scores)
    katz_centrality_scores = [0.95, 0.03, 0.65, 0.12, 0.06]
    # Relative scaling to range [1,10]
    katz_centrality_scores = [9*(i-min(katz_centrality_scores))/(max(katz_centrality_scores) - min(katz_centrality_scores)) + 1 for i in katz_centrality_scores]
    
    # Gene labels (replace with your gene labels)
    gene_labels = ["Gene1", "Gene2", "Gene3", "Gene4", "Gene5"]
    
    # Gene types (1 for oncogene, 2 for driver gene)
    gene_types = [1, 2, 1, 2, 2]
    
    # Create a graph
    G = nx.DiGraph()
    
    # Add nodes with attributes
    for i in range(len(gene_labels)):
        node_color = 'red' if gene_types[i] == 1 else 'green'
        node_shape = 'v' if gene_types[i] == 1 else 's'
        node_size = katz_centrality_scores[i]*80  # Adjust the scaling factor as needed
        G.add_node(gene_labels[i], color=node_color, shape=node_shape, size=node_size)
    
    node_colors = [v['color'] for v in dict(G.nodes(data=True)).values()]
    # Add edges from the adjacency matrix
    for i in range(len(gene_labels)):
        for j in range(len(gene_labels)):
            if adjacency_matrix[i][j] > 0:
                G.add_edge(gene_labels[i], gene_labels[j], weight=katz_centrality_scores[i], color=node_colors[i])
    
    # Extract node attributes
    node_colors = [G.nodes[n]['color'] for n in G.nodes()]
    node_shapes = [G.nodes[n]['shape'] for n in G.nodes()]
    node_sizes = [G.nodes[n]['size'] for n in G.nodes()]
    edge_colors = [G[u][v]['color'] for u, v in G.edges()]
    
    # Extract edge weights
    edge_weights = [G[u][v]['weight'] for u, v in G.edges()]
    
    # Draw the graph
    pos = nx.spring_layout(G, seed=42)  # You can use other layout algorithms
    curved_edges = [edge for edge in G.edges() if reversed(edge) in G.edges()]
    straight_edges = [edge for edge in G.edges() if not reversed(edge) in G.edges()]
    
    nx.draw(G, 
            pos, 
            node_color=node_colors, 
            node_size=node_sizes, 
            edge_color=edge_colors, 
            # node_shape=node_shapes,
            width=edge_weights, 
            with_labels=True, 
            edgelist=straight_edges, 
            arrowsize=25, 
            arrowstyle='->')
    
    nx.draw(G, 
            pos, 
            node_color=node_colors, 
            node_size=node_sizes, 
            edge_color=edge_colors, 
            # node_shape=node_shapes,
            width=edge_weights, 
            with_labels=True, 
            edgelist=curved_edges, 
            connectionstyle='arc3, rad = 0.25',
            arrowsize=25,
            arrowstyle='->')
    
    # Create a legend
    red_patch = plt.Line2D([0], [0], marker='v', color='red', label='Oncogene', markersize=10, linestyle='None')
    green_patch = plt.Line2D([0], [0], marker='s', color='green', label='Driver Gene', markersize=10, linestyle='None')
    plt.legend(handles=[red_patch, green_patch], loc='upper right')
    
    # Show the plot
    plt.title('Gene Network')
    plt.axis('off')  # Turn off axis labels and ticks
    plt.show()

运行上述代码后，图的边属性如下：

 list(G.edges(data=True))

 [('Gene1', 'Gene2', {'weight': 10.0, 'color': 'red'}),
 ('Gene1', 'Gene5', {'weight': 10.0, 'color': 'red'}),
 ('Gene2', 'Gene5', {'weight': 1.0, 'color': 'green'}),
 ('Gene3', 'Gene1', {'weight': 7.065217391304349, 'color': 'red'}),
 ('Gene3', 'Gene4', {'weight': 7.065217391304349, 'color': 'red'}),
 ('Gene4', 'Gene3', {'weight': 1.8804347826086958, 'color': 'green'})]

`

上面的代码生成以下图表（这是不正确的）：

请注意，图像不符合代码中提到的节点和边标准。例如，

1. 节点对 (Gene3, Gene4) 和 (Gene4, Gene3) 的边颜色和权重分别为 (red,7.06) 和 (green,1.88)。但在生成的图中，(Gene4, Gene3) 之间的边颜色显示为红色（不是绿色），并且边厚度与 (Gene3, Gene4) 相同（这是错误的）。

2. 当我取消注释

   node_shape

   中的

   nx.draw

   参数时，出现以下错误：

   ValueError: Unrecognized marker style ['v', 's', 'v', 's', 's']

   。我无法弄清楚如何给出每个基因类别的节点形状（例如，致癌基因为三角形，驱动基因为正方形）

任何人都可以向我建议我在上面的代码中缺少什么吗？

谢谢。

Answer 1

curved_edges = [edge for edge in G.edges() if reversed(edge) in G.edges()]
straight_edges = [edge for edge in G.edges() if not reversed(edge) in G.edges()]

这里有两个不同的列表，并调用绘制两次，每个列表一次。然而，每次调用时您都会传递相同的颜色列表

edge_color=edge_colors, 

您需要两个颜色列表，一个用于直边，一个用于曲边，以便绘制方法将为每个边获得正确的颜色。