Charged cysteine monolinks in spectral library generation

[chloe-pepper]

Summary
I'm labelling reduced and oxidised cysteines with a cleavable charged monolink (maleimide-AASK[dde]HHHHHH, light/heavy pair), enriching them together, then cleaving with hydrazine post-digestion and re-trypsinising. Samples are analysed on an Astral instrument. The positive charge on the lysine affects fragmentation patterns and fragment m/z values, so it can't be modelled as a simple mass offset. I've explored mass offset searches and labile search workflows, but neither captures the charge-dependent behaviour. Can FragPipe handle this?

The Problem
Standard spectral library formats cannot encode the charge on this modification:

• Cannot model as simple mass addition (--var-mod C,+520.xx)
• Charge affects fragment m/z values and fragmentation patterns
• DIA-NN's neural network library search will struggle to match these spectra

Vadim Demichev's workaround (convert .predicted.speclib to .tsv, manually edit fragments, re-import) exists but isn't scalable.

Questions

1. Does MSFragger support charge state specification in modifications, or is it strictly mass-based?
2. Can offset search be restricted to cysteines AND capture hydrazine cleavage diagnostic ions?
3. Are there planned enhancements to preserve charge information in spectral libraries?

[fcyu]

Does MSFragger support charge state specification in modifications, or is it strictly mass-based?

No, only support mass-based.

Can offset search be restricted to cysteines AND capture hydrazine cleavage diagnostic ions?

I will let Dan @dpolasky answer this question.

Are there planned enhancements to preserve charge information in spectral libraries?

Could you elaborate on it more? What does "preserve charge information" mean?

Thanks,

Fengchao

[dpolasky]

Hi,

1. Yes, MSFragger supports charged modifications in the sense that you can adjust the mass of the modification to account for the replacement of a proton by the intrinsic charge of the modification. We do this for metal cation adducts, for example, in the common-PTMs-mass-offset workflow. From the search engine's perspective, there is no difference between this and specifically encoding the charge, because everything is ultimately just a mass internally.

2. Yes. Mass offsets can be restricted to certain amino acids and diagnostic ions can be specified (using the labile search mode). See the tutorial here for details: https://fragpipe.nesvilab.org/docs/tutorial_labile.html.

3. I think this is not needed if the mass of the modification is adjusted to account for subtracting a proton.

Best,
Dan