FDR Filter

[herbiyuliantoro]

Hello Dr. Meyer

Thanks for the wonderful software you developed for Direct Infusion. Is it possible to change the FDR filter to 0.5 or 0.1? Sometimes, people need to know protein detection with low confidence. thank you

[jessegmeyerlab]

Thanks for your interest in our work. Although I would not recommend using a higher threshold it could easily be changed. @CCranney could point you towards what line to change in your own fork.

Caleb do you think we should add this as an option?

[CCranney]

Hi @herbiyuliantoro! Thank you for your interest. This has not been directly implemented as an end-user adjustable value, though as @jessegmeyerlab said it is quite easy to change in the code itself. I made a branch main-fdrCutoff-0p1 that changed the FDR from 0.01 to 0.1 and verified that it works on a quick manual test. In summary, you change all three instances of fdrCutoff=0.01 to fdrCutoff=0.1 in zodiaq -> scoring -> fdrCalculationFunctions.py.

@jessegmeyerlab - At present, I do worry about adding too many low-level variables that most end users will not need or want, so do not think this should be added as an end-user option. I could see it being more confusing than helpful. I'd be interested in revisiting the issue if more use cases become apparent, though.

[herbiyuliantoro]

Thank you so much for your valuable insight. Will do it!

[herbiyuliantoro]

The reason I ask like this is because the other software has this option, like Proteome Discoverer, Maxquant, and DIA-NN. However, your software is incredible. thanks!

[CCranney]

Thank you so much! We'll take a look at those internally to see if that would be a good development direction to take. In the meantime, feel free to reach out if you have any other questions or ideas!

[herbiyuliantoro]

Hello,

Sorry just let you know that

 I have an issue with processing filtering and scoring, I copied the error code below, and then I tried to fix it, and after that, it went well. I copied the modified code below of the error code.
 
 traceback (most recent call last):
File "", line 198, in _run_module_as_main
File "", line 88, in run_code
File "C:\Users\herbi\AppData\Local\Programs\Python\Python312\Scripts\zodiaq.exe_main.py", line 7, in

File "C:\Users\herbi\CsoDIAq\src\zodiaq\zodiaq.py", line 34, in main
run_scoring(args)
File "C:\Users\herbi\CsoDIAq\src\zodiaq\zodiaq.py", line 101, in run_scoring
proteinDf = create_protein_fdr_output_from_peptide_fdr_output(peptideDf)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\herbi\CsoDIAq\src\zodiaq\scoring\fdrCalculationFunctions.py", line 52, in create_protein_fdr_output_from_peptide_fdr_output
proteinDf["uniquePeptide"] = determine_if_peptides_are_unique_to_leading_protein(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\herbi\CsoDIAq\src\zodiaq\scoring\fdrCalculationFunctions.py", line 191, in determine_if_peptides_are_unique_to_leading_protein
uniquePeptides[uniqueValuesDf.index] = 1
~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^
IndexError: index 84 is out of bounds for axis 0 with size 82

=================================================================

def determine_if_peptides_are_unique_to_leading_protein(proteinDf):
"""
Determines if the peptide in the peptide column is unique to the protein
in the leadingProtein column.

Parameters
----------
proteinDf : pandas DataFrame
    The library-query match output ordered around leading proteins.

Returns
-------
uniquePeptides : list
    A list containing binary values. The length of the list corresponds to the
        dataframe provided as input. 0 indicates the peptide is NOT unique, 1 that
        it is unique.
"""
# Create an array of zeros with the same length as proteinDf
uniquePeptides = np.zeros(len(proteinDf), dtype=int)
# Group by leading protein and find groups with exactly one entry
uniqueValuesDf = proteinDf.groupby("leadingProtein").filter(lambda group: len(group) == 1)
# Set 1 for unique peptide indices, ensuring no out-of-bound index is used
for idx in uniqueValuesDf.index:
    if idx < len(uniquePeptides):
        uniquePeptides[idx] = 1
return uniquePeptides.tolist()