Question: How to plot heatmap for the matrix for the pairwise comparison of sequence similarity in R ?
0
lokraj200390 wrote:

I am working sequence alignment and I want to show the percentage of difference among all the pairs. I got a pairwise comparison table in CSV format from MEGA in the following format :

``````Table 1.

a
b   1
c   2   5
d   1   8   7
e   3   9   6   6
``````

As you can see the output table does not contain names for the columns. Ideally, columns should be a,b,c,d,e so that (a,a=0), (a,b=1),(a,c=2) and so on.

My initial thought is that it might be easier if I convert this table into the format like follow :

Table 2.

``````    a   b   c   d   e
a   0   1   2   1   3
b   1   0   5   8   9
c   2   5   0   7   6
d   1   8   7   0   6
e   3   9   6   6   0
``````

1) Is it possible to make heatmap without converting table 1 to table 2 in R ?

2) Since my table has 380 columns , how do I convert table 1 to table 2 in R ?

3) Are there any visualization tool better than heatmap to plot this kind of data in R ?

heatmap R • 2.5k views
modified 2.1 years ago by Chirag Parsania1.8k • written 2.1 years ago by lokraj200390
4
Chirag Parsania1.8k wrote:

You can use `corrplot` from package `corrplot` to generate the matrix heatmap. You can use your Table 2 as input. However, it will show you one of the half of given matrix. `as.matrix` should work to convert Table 1 to Table 2

``````corrplot::corrplot(as.matrix(dd) , is.corr = FALSE , type = "upper",diag = F,method = "color",tl.cex = 3,cl.cex = 1)
`````` 