Human–AI Co-Design and Reproducible Cheminformatic Characterisation of AIIP-RA-DMAE-001, a Rosmarinic-Acid-Derived Immunopharmacological Candidate

Rosmarinic acid is a chemically attractive immunopharmacology scaffold, but its high polarity, ionization, metabolic lability, and incomplete target definition limit translation. This original communication defines AIIP RA DMAE 001, the internal code assigned to rosmarinic acid 2-(N,N-dimethylamino)ethyl ester, an artificial-intelligence-guided phytochemical-derived analogue generated within an AI-based immunopharmacology program. The design replaces the parent carboxylic acid with an ionizable 2-(dimethylamino)ethyl ester while retaining the natural rosmarinic-acid stereocenter and E-caffeoyl geometry. The free-base structure has the formula C22H25NO8, an average molecular weight of 431.44 g mol−1, and the isomeric SMILES CN(C)CCOC(=O)[C@@H](Cc1ccc(O)c(O)c1)OC(=O)/C=C/c1ccc(O)c(O)c1. Public exact-identity searches performed on 10 October 2026 found no individualized record for this structure, although closely related alkyl rosmarinates, methyl-rosmarinate derivatives, and a permanently charged rosmarinylcholine analogue are known and broad patent Markush language may encompass aminoalkyl esters. The compound is therefore described as putatively structurally novel, not as pharmacologically or therapeutically novel. A prior design record reported hypothesis-generating AutoDock Vina scores of −7.8, −8.1, and −8.3 kcal mol−1 against generic IgG, IgA, and IgM structures, respectively; these values do not establish binding and require prospective reproduction. The tertiary amine is expected to permit salt formation and substantial protonation near physiological pH, whereas the catechols and ester linkages create oxidation, assay-interference, hydrolysis, and distribution liabilities. We propose an evidence-gated program spanning synthesis and spectroscopic confirmation, pKa and stability measurement, orthogonal target-engagement studies, innate and adaptive immune phenotyping, and exposure-linked disease models. The most credible initial translational hypotheses concern type I interferon signaling and macrophage inflammatory state; direct immunoglobulin modulation should remain a higher-bar secondary hypothesis. AIIP RA DMAE 001 is thus a testable molecular proposal and a framework for transparent progression from AI design to immunological evidence.