Canonical Pairing Without Separation: Dual Tethers, Exact Registry Closure, and the Strand-Coding Gate

Abstract

A stack of six substituted chrysene units admits twenty classes under the symmetries available to it, and the filing on which the construction rests gives two tethers from the chrysene: one to adenosine at its C2, one to cytosine at its C5. A predecessor uses the first and reads the carrier against four coding bases under pairing that is in three cases non-canonical. This paper uses both, reads it under full Watson–Crick pairing, and does not separate the nucleotides from the chrysene at any stage.

The registry closes exactly. Two of the twenty classes have a constant agreement sequence — the all-agree and the all-disagree extremes — and they are the two tether patterns, so they are consumed rather than chosen. The eighteen remaining meet the eighteen slots with a perfect matching and no spare class, and at one slot the assignment is forced outright. No codon is lost by the consumption: a class seeds a codon and propagation fills its slot.

The heteroatom gate is made rather than inherited. The adenosine tether attaches through an ether at a position on the pairing edge, where an unmodified pair makes a weak third contact; the cytosine tether attaches in the major groove and disturbs nothing. Both tethers perturb the pairs they sit on, so a condition turning on which pair is disturbed does not discriminate between them. What does are two invariants the construction supplies and its predecessor cannot: which strand carries the tethered base, and which member of the pair codes. The first is the predecessor's own gate, derived here from a construction in which no pair is defective; over twenty-two published tables it covers one hundred and ten of one hundred and thirty-two nitrogen-bearing structures, missing one structure in every table and always the same one. The second covers all one hundred and thirty-two. Their conjunction is necessary at five structures of six and sufficient at five of the six it admits.

Three results are negative and are not caveats. Exact closure leaves one hundred and ninety-two admissible assignments, so the selection problem of the predecessor is narrowed nineteenfold and survives. It does not force the slot set: more than half the slot sets of the right shape close exactly, so the division of the blocks is a choice. And the gate admits seventeen structures of twenty, which is less selective than the eleven of the gate it replaces; a strictly tighter condition exists, necessary on the same six structures, that the construction does not explain. Two of the three were found by running computations the draft had registered as targets. The Watson–Crick partition, which that paper recovers from block structure and reports as its strongest evidence, is here assumed as the tether rule: this paper explains what its predecessor observed, and the explanation costs the observation. Neither reading is derived from the other and the filing does not say which is physical. The two are set side by side and not ranked.