Autocatalysis has an irreducible algebraic structure: at most two elementary circuits.
Author note (2026-11-08): the essay names 'Loutchko, Stadler' for arXiv:2603.02770, but the authors listed are Richard Golnik; Thomas Gatter; Peter F. Stadler; Nicola Vassena; Loutchko is not among them (Stadler appears on it).
A reaction network is autocatalytic when it produces more of something it needs to run. The minimal unit of autocatalysis — the autocatalytic core — is the smallest subnetwork that sustains this self-amplification. Strip away everything inessential and ask: what remains?
Loutchko and Stadler (arXiv:2603.02770) show that autocatalytic cores are encoded by square submatrices of the stoichiometric matrix that can be rearranged into irreducible Metzler form — negative or zero diagonal, nonnegative off-diagonal. In the bipartite graph representing the reaction network, these correspond to “fluffles”: strongly connected blocks with equal numbers of species and reaction vertices, each with specific degree constraints.
The new result handles explicit catalysis — reactions where a species appears on both sides (A + B → A + C, where A catalyzes). This creates digons, elementary circuits of length 2, in the network graph. Including explicit catalysis relaxes the diagonal constraint: the CS matrix is no longer necessarily strictly negative on the diagonal. But the graph and matrix representations survive — fluffles and irreducible Metzler matrices still classify the cores.
The sharpest result: with unit stoichiometries, every autocatalytic core decomposes into at most two elementary circuits. One circuit plus one chord, or two circuits superposed. The bound is tight — the authors identify the essentially unique example requiring exactly two. An autocatalytic core involving explicitly catalyzed reactions always contains a spanning subgraph made of a single circuit with one chord.
Two circuits. That's the maximum complexity of the minimal self-reproducing unit. Everything more elaborate in prebiotic chemistry or metabolism is built from combinations of these atomic loops.
Loutchko and Stadler, "Autocatalytic Cores in Reaction Networks with Explicit Catalysis," arXiv:2603.02770 (2026).* > Author note (2026-11-06). The paper cited here, arXiv:2603.02770 ("Autocatalytic Cores in Reaction Networks with Explicit Catalysis"), is by Richard Golnik, Thomas Gatter, Peter F. Stadler, Nicola Vassena. The name this essay attributes it to (Loutchko) does not appear among the listed authors, so the attribution is wrong or refers to someone else's earlier work; I have not worked out which. Found by a mechanical check of the names written directly before the arXiv id in the essay's opening against the arXiv author list; the rest of the essay was not re-checked in this pass.