Skip to content

Resolving power & FWHM

Every resolvability verdict in ProteoformTracker — whether two charge-state peaks are "clean" or "overlapping" in the MS1 stats panel, whether a fragment ion is distinguishable, whether a confounding protein's peak collides with your target's — comes back to the same instrument model.

The Orbitrap resolving-power law

Resolving power is quoted at one reference m/z (conventionally 200) and degrades as m/z increases:

\[ R(m/z) = R_{ref} \times \sqrt{\frac{m/z_{ref}}{m/z}} \]

This converts to a mass-domain peak width (full width at half maximum):

\[ \Delta M_{FWHM} \text{ (Da)} \approx \frac{M}{R(m/z_{\text{best}})} \]

where M is the ion's mass and m/z_best is whichever charge state in its envelope gives the smallest (best) ΔM_FWHM.

Worked example

A 16 kDa proteoform at charge state z=20 has m/z ≈ (16,000 + 20 × 1.007) / 20 ≈ 801. At R_ref = 120,000 and m/z_ref = 200:

R(801) = 120,000 × sqrt(200 / 801) ≈ 60,000
ΔM_FWHM ≈ 16,000 / 60,000 ≈ 0.27 Da

So at that charge state, two proteoforms of this protein would need to differ by more than roughly 0.27 Da (times the safety margin, below) to show up as two separate peaks rather than one blurred one.

Safety margin

Being theoretically resolvable and being confidently, practically separable in a real spectrum aren't quite the same thing. ProteoformTracker applies a safety margin (default 1.75× ΔM_FWHM, adjustable in MS resolution parameters) before calling a pair "clean" rather than merely "borderline" — the design spec's own recommendation is a 1.5–2× margin.

Where this shows up in the UI

  • MS1 charge-envelope peaks stats tile (Section 1 and Section 2): counts how many of the checked/target proteoform's charge-state peaks are "clean" (no other proteoform/confounder within the safety margin) vs. "overlapping."
  • Hovering an MS1 peak: the tooltip reports the nearest other peak's distance in units of the instrument's own FWHM at that m/z — under 1× reads as "effectively the same detected peak," a few × as "partially separated," many × as "clearly resolved."
  • Confounder search: a charge-state peak from a candidate protein counts as colliding with your target's peak using this exact same tolerance.

What this model does not capture

This is an Orbitrap-specific scaling law — FT-ICR and other platforms have a different resolving-power relationship and would need a different formula. It's also a peak-width model, not a full simulation of isotope fine structure; see Charge-state envelopes for how the actual isotope pattern (or its smooth envelope, when unresolved) is predicted.