search

Noble-Lab / FDRBench

Estimation of false discovery proportion for mass spectrometry proteomics data using entrapment
Apache License 2.0
4 stars 1 forks source link

No FDP values for a dataset of 400,000 precursors #3

Open chinmayaNK22 opened 1 month ago

chinmayaNK22 commented 1 month ago

Hi,

The FDP calculation at the precursor level for from a proteogenomic search consisting >400,000 peptide precursor is resulting with only zero values. I am not able to figure out the issue here. I tried to do the same with precursors with 0.01 < q-values > e7, which comprises just over 250,000 peptide precursors. Still the resulting FDP is zero for all the peptides.

Could you please tell me why it is so, is there are issue with the input file. However, I have not faced any major errors or warnings.

The peptide_precursor.tsv and entrapment_pep.txt files are available in the below OneDrive link for your reference. Let me know if you need more information.

https://qubstudentcloud-my.sharepoint.com/:u:/g/personal/3058937_ads_qub_ac_uk/ESy8UHLFdyxJpfJuB0PyySEBVO5ybStIWE8t7AzLFzFvtg?e=YhFTx6

wenbostar commented 1 month ago

Hi Chinmaya,

Could you please share the FDRBench command line with me?

Bo

chinmayaNK22 commented 4 weeks ago

Hi Bo,

Here is the command line used for executing the precursor level FDP

>java -jar fdrbench-0.0.1.jar -i peptide_precursor.tsv -fold 1 -pep entrapment_pep.txt -level precursor -o peptide_precursor_FDP.csv -score "score:0"

Are you expecting that the command line or arguments used causing any issue here?

Chinmaya

chinmayaNK22 commented 2 weeks ago

Hi Bo,

As I tried to execute the FDRBench with smaller dataset, I faced the similar problem.

Here, I am using my python script to change the format of the Spectronaut exported precursors list to FDRBench compatible input format. The python script mainly calculates the score by sorting EG.Cscore and adds this new column to the file.

In the meantime, I also created the same input file by manually sorting the file in excel and submitted that as input.

Surprisingly, the providing excel formatted input file gives the an output with FDP values from FDRBench, whereas the output generated for the python script formatted input file results only with zeros as FDP.

I have attached the input file in both the formats. Does the issue lies with the FDRBench in recognizing python output correctly or is there any issues with the python script I generated?

I have shared everything here for your reference.

https://qubstudentcloud-my.sharepoint.com/:u:/g/personal/3058937_ads_qub_ac_uk/EVvxLwwVWNdBhQPegu7dwM4B7lvxeKk8dleVVHngo4C8Ww?e=RhZoy0

Chinmaya