Technical Inquiry on DDA Quantification of Adjacent Phosphorylation Sites with Overlapping RT and Ion Mobility

[2021221221]

Dear Fragpipe team,

I am exploring Fragpipe's capability to distinguish and quantify adjacent phosphorylation sites on the same peptide in DDA mode, particularly when their retention time (RT) and ion mobility are highly similar. Could you clarify the following points regarding the underlying methodology?

MS/MS Spectral Specificity:
How does Fragpipe (e.g., MSFragger/Philosopher) ensure accurate quantification of adjacent phosphosites when their precursor ions co-elute and share near-identical RT/ion mobility? Does the pipeline prioritize site-determining ions (e.g., specific b/y ions with phosphorylation modifications) in MS/MS spectra to resolve such cases?

Localization Confidence and Quant Integration:
For peptides with ambiguous site localization (e.g., adjacent Ser/Thr residues), how are quantitative values assigned to each site? Does PTMProphet’s localization probability directly influence the quantitative output (e.g., by weighting intensities based on site confidence)?

Ion Mobility and Co-elution Challenges:
In cases where ion mobility separation (e.g., TIMS) fails to resolve isomers, does Fragpipe leverage MS1-level features (e.g., extracted ion chromatograms, isotopic patterns) or advanced algorithms to deconvolute contributions from co-eluting phosphoforms?

Best Practices for Ambiguous Sites:
Are there recommended parameters or post-processing filters (e.g., Ascore thresholds, intensity correlation checks) to minimize quantitative cross-talk between adjacent sites?

Thank you for your time.

Best regards

Jie Yu

[dpolasky]

Hi Jie Yu,

How does Fragpipe (e.g., MSFragger/Philosopher) ensure accurate quantification of adjacent phosphosites when their precursor ions co-elute and share near-identical RT/ion mobility? Does the pipeline prioritize site-determining ions (e.g., specific b/y ions with phosphorylation modifications) in MS/MS spectra to resolve such cases?

PTMProphet uses site-determining ions in the MS/MS spectra to determine the localization, so we recommend using it for phospho analyses where adjacent sites are a concern. You can see the PTMProphet paper for all the details: https://pubs.acs.org/doi/10.1021/acs.jproteome.9b00205

Localization Confidence and Quant Integration:
For peptides with ambiguous site localization (e.g., adjacent Ser/Thr residues), how are quantitative values assigned to each site? Does PTMProphet’s localization probability directly influence the quantitative output (e.g., by weighting intensities based on site confidence)?

There is a localization probability filter on both the label free and isobaric labeling Quant tabs, so you can choose a minimum localization probability (from PTMProphet) to include sites in the quantitation. Ambiguous localizations (i.e., those below the threshold) are not included in the quantitation.

Ion Mobility and Co-elution Challenges:
In cases where ion mobility separation (e.g., TIMS) fails to resolve isomers, does Fragpipe leverage MS1-level features (e.g., extracted ion chromatograms, isotopic patterns) or advanced algorithms to deconvolute contributions from co-eluting phosphoforms?

No, coeluting isomers that generate only a single MS2 spectrum will be assigned only one "best" localization. If there are multiple MS2 spectra, multiple localizations can be assigned, but this is unlikely for coeluting/unseparated isomers due to dynamic exclusion.

Best Practices for Ambiguous Sites:
Are there recommended parameters or post-processing filters (e.g., Ascore thresholds, intensity correlation checks) to minimize quantitative cross-talk between adjacent sites?

The localization probability threshold for quant is the most straightforward method. If there are particular sites of interest, their spectra can always be checked and confirmed using the FragPipe-PDV viewer. But this is a difficult problem and I do not think anyone has a perfect solution.

Best,
Dan