首页 › R语言论坛 › pheatmap实用参数（二）

天涯随风 發表於 2023-3-12 23:54:00

pheatmap实用参数（二）

<pre><code># Create test matrix（造数据）
set.seed(6)
test = matrix(rnorm(200), 20, 10)
test = test + 3
test = test + 2
test = test + 4
colnames(test) = paste("Test", 1:10, sep = "")
rownames(test) = paste("Gene", 1:20, sep = "")
</code></pre>
<h2 id="正文从这里开始"><em>正文从这里开始</em></h2>
<pre><code># Show text within cells
pheatmap(test, display_numbers = TRUE)
pheatmap(test, display_numbers = TRUE, number_format = "\\%.1e")
pheatmap(test, display_numbers = TRUE, display_numbers = matrix(ifelse(test &gt; 5, "*", ""), nrow(test)))
</code></pre>
<ul>
<li>
<p>display_numbers = T 即Show text within cells</p>
</li>
<li>
<p>number_format 可以格式输出display_number</p>
</li>
<li>
<p>或者干脆自定义一个matrix通过display_numbers参数进行display</p>
</li>
</ul>

<pre><code># legend（图例）的设置选项
pheatmap(test)# p1
pheatmap(test, legend_breaks = -1:7)# p2
pheatmap(test, legend_breaks = 1:6, legend_labels = c("6","6", "6", "6", "6", "6"))# p3
pheatmap(test, legend_breaks = 9:14, legend_labels = c("6","6", "6", "6", "6", "6"))# p4
</code></pre>
<ul>
<li>
<p>pheatmap函数会在内部算出legend的数值范围，本例中大概是 -1:7</p>
</li>
<li>
<p>在数值范围内，我们可以设定legend_breaks，以及对legend_breaks这个label的文本展示</p>
</li>
<li>
<p>legend_breaks和legend_labels是有一个对应关系的，否则报错如下</p>
<p>Error in pheatmap(test, legend_breaks = 1:6, legend_labels =<br>
c(“6”, “6”, : Lengths of legend_breaks and legend_labels must be<br>
the same</p>
</li>
<li>
<p>假如我们的设定范围超出，就如p4所示</p>
</li>
</ul>
<p><img src="https://img2023.cnblogs.com/blog/1609640/202303/1609640-20230312234145782-1888955972.png"></p>
<pre><code># Fix cell sizes and save to file with correct size
pheatmap(test, cellwidth = 15, cellheight = 12, main = "Example heatmap")# the title of the plot
pheatmap(test, cellwidth = 15, cellheight = 12, fontsize = 8, filename = "test.pdf")# save to pdf

# Change angle of text in the columns (0, 45, 90, 270 and 315)
pheatmap(test, angle_col = "45")
pheatmap(test, angle_col = "0")
</code></pre>
<h3 id="有关annotation"><em>有关annotation</em></h3>
<pre><code># Generate annotations for rows and columns（先造数据）
annotation_col = data.frame(
                  CellType = factor(rep(c("CT1", "CT2"), 5)),
                  Time = 1:5
                )
rownames(annotation_col) = paste("Test", 1:10, sep = "")

annotation_row = data.frame(
                  GeneClass = factor(rep(c("Path1", "Path2", "Path3"), c(10, 4, 6)))
                )
rownames(annotation_row) = paste("Gene", 1:20, sep = "")

# Display row and color annotations
pheatmap(test, annotation_col = annotation_col, cluster_cols = F)

pheatmap(test, annotation_col = annotation_col, annotation_legend = FALSE, cluster_cols = F)

pheatmap(test, annotation_row = annotation_row, cluster_rows = F)
</code></pre>
<p><img src="https://img2023.cnblogs.com/blog/1609640/202303/1609640-20230312234611984-536833328.png"></p>
<ul>
<li>
<p>annotation数据首先要组织成dataframe，dataframe中的rownames要和注释项进行对应，<br>
而column就是要展示的注释条，每个column都会生成一个注释条来显示</p>
</li>
<li>
<p>annotation_col/row 默认会有legend配合展示，annotation_legend = FALSE可以去掉legend</p>
</li>
</ul>

<pre><code># Specify colors （自定义注释条颜色）
ann_colors = list(
    Time = c("white", "firebrick"),
    CellType = c(CT1 = "#1B9E77", CT2 = "#D95F02"),
    GeneClass = c(Path1 = "#7570B3", Path2 = "#E7298A", Path3 = "#66A61E")
)

pheatmap(test, annotation_col = annotation_col, annotation_colors = ann_colors)

pheatmap(test, annotation_col = annotation_col, annotation_row = annotation_row, annotation_colors = ann_colors)
</code></pre>
<p><img src="https://img2023.cnblogs.com/blog/1609640/202303/1609640-20230312234712169-1492161202.png"></p>
<ul>
<li>注释条颜色数据要组织成list（list的灵活性就在此处凸显出来了），<br>
list中的每个元素名作为对应项（对应前述dataframe中的colnames），<br>
同时可以进一步为attribute指定颜色，例如<br>
<code>CT1 = "#1B9E77", CT2 = "#D95F02"</code></li>
</ul>
<hr>
<hr>
<h2 id="参考资料"><em>参考资料</em></h2>
<p><strong>pheatmap帮助文档</strong></p><br><br>
来源：https://www.cnblogs.com/yknNewbie/p/17209809.html

查看完整版本: pheatmap实用参数（二）