Plotting a matrix in python

Question

I have a pandas dataframe which looks like:

gene1   gene2        pvalue        qvalue
TP53.   FUBP1       6.381282e-09.  2.435575e-08
TP53.    CIC        1.570489e-22.  7.055298e-22
IDH1.    NF1        1.946551e-05.  7.116265e-05

I want a matrix as shown below:

So I to plot a matrix which has gene names as rows and columns and colours only those cells for which gene pairs are found in the pandas dataframe above. How can I achieve this in pandas.

That looks like a diagonal correlation matrix to me: seaborn.pydata.org/examples/many_pairwise_correlations.html medium.com/@szabo.bibor/… — albert
– albert, Commented Sep 12, 2021 at 11:04
That's what I thought too. But it uses .corr() function. I don't want to do that. I just want to go color those cells for which pairs are found in the pandas dataframe and color them according to pvalue column. Do you have ideas how can I modify the code to do that? — John
– John, Commented Sep 12, 2021 at 11:32
You could pre-process your data to fit a diagonal form and then plot it in a heatmap manually: stackoverflow.com/questions/65670428/… — albert
– albert, Commented Sep 12, 2021 at 11:39

mozway · Accepted Answer · 2021-09-12 12:26:42Z

1

Here is a working version:

import seaborn as sns

# cleanup
df['pvalue'] = df['pvalue'].str[:-1].astype(float)
df['gene1']  = df['gene1'].str[:-1]

idx = sorted(set(df['gene1']).union(df['gene2']))
df2 = (df.pivot(index='gene1', columns='gene2', values='pvalue')
         .reindex(index=idx, columns=idx)
      )

mask = df2.isna()&np.isnan(df2.T.values) # keep track of NAs on both combinations of genes (A/B and B/A)

df2 = df2.fillna(0)
df2 += df2.values.T  # fill matrix A/B → B/A
df2 = df2.mask(mask) # restore NAs

cmap = sns.diverging_palette(323, 101, s=60, as_cmap=True)
ax = sns.heatmap(data=-np.log10(df2), mask=np.triu(df2.values), center=0, cmap=cmap)
ax.invert_yaxis()
ax.xaxis.tick_top()

output:

answered Sep 12, 2021 at 12:26

mozway

267k13 gold badges56 silver badges106 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

Plotting a matrix in python

1 Answer 1

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related