Year: 2026
Type: preprint
Abstract: Selecting a set of non-isobaric amino acids is the standard way to guarantee unambiguous readout when peptide sequences are used to carry information. We show it is insufficient. Because tandem MS reads residues as differences between consecutive fragment masses, a single missing b/y ion merges two adjacent residues into one observed gap; if that merged gap aliases a single residue mass, the decoder does not register a dropout but emits one confident, wrong residue. Non-isobaricity cannot prevent this, because the colliding quantity is a sum rather than a residue. We formalize the missing condition—gap-merge safety, |m(a) +m(b)−m(c)| > τ for all a, b, c in the alphabet—as a decidable predicate with a machine-checked soundness proof in Lean 4, and characterize the collision structure of the proteinogenic alphabet exhaustively: beyond the Leu/Ile isobar and the Gln/Lys near-isobar, exactly two summed-gap collisions exist, Gly + Gly = Asn and Gly + Ala = Gln, making glycine disproportionately expensive. In simulation, alphabets satisfying gap-merge safety produce zero silent errors at every fragment dropout rate from 0 to 0.4, where alphabets chosen by non-isobaricity alone produce 1.0% to 6.9%; every failure becomes a detectable erasure rather than a silent substitution.
Source: ChemRxiv
Author: Laalini Bhogadi
Institution: Indian Institute of Management Kozhikode
Language: English
Subfield: Spectroscopy
Field: Chemistry
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7211991001