Glossary

Proteoform : The shared internal representation every input path resolves into before any scoring runs: a mature amino-acid sequence (after N-terminal Met excision and signal-peptide/propeptide removal) plus a list of applied PTMs.

Top-down proteomics (TDP) : Mass spectrometry performed on intact proteins, without proteolytic digestion.

Middle-down proteomics (MDP) : A limited (partial) protease digestion first, producing large peptides, analyzed the same way top-down would analyze an intact protein. See Middle-down mode.

Resolving power R(m/z) : An Orbitrap's mass-resolving capability, quoted at a reference m/z and decreasing as m/z increases. See Resolving power & FWHM.

ΔM_FWHM : The mass-domain peak width (full width at half maximum) implied by resolving power at a given m/z — the practical yardstick for "how close is too close to resolve."

Safety margin : A multiplier (default 1.75×) applied over ΔM_FWHM before calling a pair confidently — not just theoretically — separable.

Charge-state envelope : The distribution of charge states (and, within each, the isotope pattern) a proteoform is predicted to produce. See Charge-state envelopes.

Averagine : A closed-form statistical approximation to a protein's isotope pattern, used instead of full combinatorial isotopologue enumeration (intractable at intact-protein scale).

Fragmentation propensity : A per-bond score estimating how likely a given backbone bond is to fragment under HCD/CID, built from residue-pair, positional, length, and PTM-proximity effects — not an intensity prediction. See Fragmentation propensity & scoring modes.

b-ion / y-ion : The two terminal fragment-ion series modeled (N-terminal and C-terminal fragments, respectively) — the only ion types modeled; internal fragments and ETD/ECD c/z ions are out of scope for now.

Relevant proteoform : A proteoform you explicitly checked for comparison (an isoform, a PTM variant, a novel sequence, or an rMATS-matched transcript).

Confounding protein : A protein from the reference proteome — unrelated to your target by gene or sequence — that shares intact mass and/or a charge-state m/z peak with your target. See Relevant vs. confounding proteins.

found_via : How a confounder candidate was found: mass (intact-mass window), mz (charge-state peak collision), or both.

Tier (unique / partial / common / neutral) : How many of the other proteoforms/confounders in the current comparison share a given exon block or fragment ion. See Tier colors.

Calibrated (GLM) scoring mode : The primary fragmentation-propensity scoring formula, length-aware and calibrated against real matched b/y ions from published top-down datasets.

RF ranking scoring mode : A length-free, random-forest-based relative ranking, for long proteoforms whose Calibrated-mode scores are all suppressed by their own length. Not comparable in magnitude to Calibrated mode.

PTM spec : The text-box syntax for specifying modifications on a proteoform, e.g. 133_S_Phospho. See PTM specification syntax.

rMATS : A widely-used tool for detecting differential alternative-splicing events (SE, MXE, A3SS, A5SS, RI) from RNA-seq data. See Option 3.

SE / MXE : rMATS event types supported by Option 3 — Skipped-Exon and Mutually-eXclusive-Exons, respectively.

IsoPepTracker : ProteoformTracker's bottom-up companion project, at github.com/HuangLabAtUAB/IsoPepTracker.