how can i filter the results of fp

[hima111997]

how can i filter the results of interactions based on their proportion to enhance the plot?

[cbouy]

Hi,

You can directly filter the pandas DataFrame output by fp.to_dataframe and input that to the Barcode class (which is run under the hood when calling fp.plot_barcode).

Here's an example:

from prolif.plotting.barcode import Barcode

# example: filter interaction that occur in less than 30% of frames
threshold = 0.3
df = fp.to_dataframe()
# only keep interactions occurring more frequently than the threshold
above_threshold_mask = df.mean() > threshold
df_filtered = df.loc[:, above_threshold_mask]
# plot
Barcode(df_filtered).display()

Feel free to reopen the issue if you have another related question

[Mishakolok]

Hello, I have tried to use the suggested solution, however, I get an error "KeyError: 'Level ligand not found'"
I think this happens because the df_filtered DataFrame does not have the first row with the name of the ligand. However, I don't have enough experience with Pandas to fix it straightaway.