如何使Matplotlib imshow与dna_features_viewer的X轴物理刻度对齐？

问题根源 #

imshow默认启用aspect='equal'，强制图像宽高比一致，导致子图宽度被限制，无法和下方DNA Features Viewer生成的子图X轴物理长度匹配；同时两者的X轴索引范围也存在细微差异。

修复步骤 #

1. 给imshow添加aspect='auto'参数，让图像自适应子图宽度
2. 调整ax0的X轴范围，与DNA Features Viewer的Genbank索引（1到850）对齐

修改后的完整代码 #

from dna_features_viewer import GraphicFeature, GraphicRecord
import matplotlib
from matplotlib import rcParams
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd


rcParams["figure.figsize"] = 15, 12
fig, (ax0, ax1) = plt.subplots(2, 1, height_ratios=[5, 1], sharex=True)

# ax0 - 修改部分：添加aspect='auto'，设置xlim匹配genbank索引
data_heatmap = np.random.rand(850, 850)
# 关键：添加aspect='auto'取消等宽高比限制
heatmap = ax0.imshow(data_heatmap, cmap="bone", aspect='auto')
fig.colorbar(heatmap, ax=ax0)
ax0.set_xlabel("Scored Residue")
ax0.set_ylabel("Aligned Residue")
ax0.set_ylim(len(data_heatmap) + 1, 1)
# 调整X轴范围，对应1-850的genbank索引（imshow默认是0-849，加0.5偏移匹配）
ax0.set_xlim(0.5, 850.5)

# ax1
domains = pd.DataFrame({"domain": ["dom1", "dom2", "dom3"],
                        "start": [9, 516, 714],
                        "end": [459, 689, 850],
                        "color": ["#0000b6", "#02eded", "#ab0000"]})
features = []
for _, row in domains.iterrows():
    features.append(GraphicFeature(start=row["start"], end=row["end"], strand=+1, color=row["color"],
                                   label=row["domain"]))
record = GraphicRecord(sequence_length=row["end"] + 1, features=features, plots_indexing="genbank")
record.plot(ax=ax1)

plt.tight_layout()
plt.show()

原理说明 #

• aspect='auto'让imshow图像不再强制等宽高，而是填充整个子图宽度，和下方子图的X轴物理长度保持一致
• 设置xlim(0.5, 850.5)是因为imshow的每个像素中心对应整数坐标，这样调整后，X轴的1-850刻度会和DNA Features Viewer的序列位置精准对齐

内容的提问来源于stack exchange，提问作者Mesmer