Chemical Identity Reconciliation and Public Database Assessment of a Fisetin Aminoethyl Ether

1. Introduction

Fisetin is a flavonol bearing hydroxyl groups at positions 3, 7, 3′ and 4′ [1]. Its distinct enolic and phenolic substitution sites permit several molecular modifications, each requiring an explicit attachment-site definition. Substituting a hydroxyl group can change hydrogen bonding, ionisation and metabolic susceptibility while retaining much of the parent scaffold. These changes provide a design rationale, but cannot by themselves establish useful exposure or pharmacological selectivity.

Fisetin derivative design is established. Chiruta and colleagues studied fisetin-derived scaffolds bearing B-ring amino groups [2]. Aminoalkyl-flavonoid programmes also predate the present proposal [3,4]. Further searches identified a 2012 series of 7-O-aminoethyl luteolin derivatives, a 2015 quercetin 3-O-methyl ether programme and an earlier study of 3,7-substituted catechol flavones [5,6,7]. A mitochondria-targeted fisetin programme introduced a butyltriphenylphosphonium substituent at position 7 with hydroxyl protection [8]. The registered aminoethyl flavone comparator has the proposed side-chain motif but lacks three hydroxyl groups [9]. PubChem CID 141343708 provides a closer scaffold precedent: a fisetin 7-O-aminoethyl linkage terminating in a substituted piperazine–fluoroquinolone group [10]. These precedents make exact attachment site, hydroxyl pattern and terminal amine identity central to the present assessment.

The present candidate originates in the author-held book manuscript Artificial Intelligence (AI)-Based Immunopharmacology: Novel Phytochemicals for the Treatment of Auto-immune Diseases [11]. Its final fisetin passage names a 7-O-(2-(dimethylamino)ethyl) ether and places it in a proposed antibody-interaction context. The chemical entries and AI-generated illustration require independent evaluation. The question addressed here is whether that named proposal can be converted into a coherent molecular specification and distinguished from relevant prior disclosures.

Tool-augmented AI systems illustrate how language-based reasoning can be linked to chemical software and experimental infrastructure [12,13]. Their performance does not validate a separate molecule nominated in an AI-assisted narrative. This communication therefore separates the source proposal, deterministic chemical results and accessible prior-art evidence. It presents an original computational audit of one nominated derivative, without introducing a trained AI model or claiming autonomous drug discovery.

2. Materials and Methods

Source Record and Study Design

The study comprised retrospective inspection of the author-supplied book, explicit molecular reconstruction, comparative descriptor calculation and a targeted public evidence assessment. The original calculations and the extended database searches are dated 12 October 2026 UTC. The chemically specific source passage occurs on digital page 318 of the 323-page copy, corresponding to printed page 304, immediately after the final caption headed Figure 36 [11]. Figure numbering and immunoglobulin labelling vary within that source; the named connectivity, rather than an illustrated ligand pose, defined the molecule analysed here.

The source caption attributes its image to an AI system prompted by the human editor. A fixed historical model, complete prompts, sampling settings and reproducible candidate-selection record are not retained. The illustration was not treated as a crystallographic measurement, a reproduced docking calculation or evidence of affinity. The present AI assistance involved literature organisation, code drafting and manuscript preparation. All reported chemical values were generated by explicit algorithms. No human participants, animals, biological specimens or newly measured experimental data were involved.

Molecular Reconstruction and Identity Checks

The curated fisetin graph was obtained from ChEBI:42567 [1]. The parent InChIKey, XHEFDIBZLJXQHF-UHFFFAOYSA-N, was used as an identity check. Only the hydrogen of the 7-OH group was replaced by CH2CH2N(CH3)2. The flavonol C2–C3 double bond and the free 3, 3′ and 4′ hydroxyl groups were retained. This produces an aryl aminoethyl ether; no carbonate or carboxyl ester is introduced.

RDKit 2025.09.6 sanitised the graphs and generated canonical isomeric SMILES, Standard InChI and InChIKey [14,15]. The candidate was encoded as a neutral tertiary-amine free base and, separately, as an amine-protonated cation. No counterion, solvate, CAS Registry Number or experimentally isolated form was assigned. The candidate's canonical SMILES is:

CN(C)CCOc1ccc2c(=O)c(O)c(-c3ccc(O)c(O)c3)oc2c1

The registered 7-aminoethyl flavone comparator was encoded independently. Its computed InChIKey was checked against the NIH NCATS record [9]. These parent and comparator checks confirm the representations of known structures; they do not establish novelty or experimental identity of the candidate.

Candidate Coding

The chemically specified structure was assigned the internal project identifier AIIP-FIS-7O-DMAE-001. AIIP denotes the author’s AI-based immunopharmacology project, FIS the fisetin scaffold, 7O the oxygen attachment site, DMAE the 2-(dimethylamino)ethyl group, and 001 this entry. The code is an organisational identifier. It is not a CAS Registry Number, a PubChem CID or evidence of a documented historical AI-generation process. Its evidence status is a computational candidate with provisional exact-structure novelty.

Descriptor Calculation

The three deliberately selected structures were fisetin, the proposed derivative and the registered aminoethyl flavone comparator. Formula, average molecular weight, monoisotopic mass, topological polar surface area, hydrogen-bond donor and acceptor counts, and Crippen cLogP were computed using RDKit. Polar surface area and lipophilicity used established fragment-based implementations [16,17]. Values describe the encoded neutral structures and are not measurements of solubility, pKa, permeability or logD at a physiological pH.

The expected singly protonated molecular-ion mass was calculated by adding 1.00727647 Da to the neutral monoisotopic mass. This supplies an analytical expectation without representing an observed spectrum. Descriptor differences were calculated arithmetically. No inferential statistics or significance tests were applied to the selected three-structure comparison. The chemical script, inputs, outputs and two-dimensional representations accompany the manuscript.

Public Database and Literature Searches

The extended search used the neutral and amine-protonated InChIKeys, the explicit neutral SMILES, C19H19NO6, the named fisetin ether and aminoethyl/dimethylaminoethoxy variants. PubChem PUG-REST queries included direct InChIKey retrieval, SMILES identity searches with same_connectivity and same_tautomer, a protonated-SMILES identity search, substructure searching capped at 100 records and a molecular-formula query capped at 1000 records [18]. Successful fisetin and aminoethyl-flavone retrievals served as positive controls. Every returned substructure record was checked against the candidate using its retrieved formula and InChIKey; substructure matching was not equated with exact identity.

ChEBI’s public REST text endpoint was queried with both candidate keys and the parent key; its formula query was also inspected [19]. ChEMBL’s public molecule endpoint was filtered by each candidate Standard InChIKey and by the full molecular formula, with a fisetin control [20]. Negative query responses were distinguished from connection failures and authorisation errors. Public database records do not comprehensively index every published compound, salt, drawing or patent genus.

Native PubMed and Europe PMC searches used fisetin combined with dimethylamino, aminoethyl and aminoalkyl terms. Europe PMC’s two returned pages were inspected at the title and metadata level; selected original research records were then followed by title, DOI and citation links. Search-engine discovery also used the exact candidate key, name variants and Chinese terms. The full 2012 aminoethyl-flavone paper was obtained and its Figure 1, Scheme 1, Table 1 and named final products were inspected [5]. Accessible full text supported the 2024 mF7 comparison [8]. The 2000, 2015, 2016, 2020 and 2023 programmes remained partly represented by abstracts or metadata during this extension [3,4,6,7,21].

Public patent text inspection covered WO2008021210A2, WO2009158031A2, US20090082400A1 and WO2013025498A1 [22,23,24,25]. Generic substituent definitions were recorded separately from named examples. CAS Common Chemistry’s attempted API query returned an authorisation error; direct Espacenet and PATENTSCOPE retrievals of the selected WO2013025498 record returned access errors. These attempts supplied no absence evidence. Licensed CAS SciFinder-n and Reaxys structure/substructure searches, exhaustive supplementary-drawing review and specialist Markush mapping were not completed. No legal determination of claim coverage or patentability was made. This was a targeted evidence assessment, not a systematic review; the query counts are not PRISMA counts. The additional Search Evidence archive records the completed API queries and their limits.

3. Results

Reconciled Molecular Identity

The named transformation changes the elemental composition according to C15H10O6 − H + C4H10N = C19H19NO6. The resulting neutral candidate has molecular weight 357.362 g/mol, monoisotopic mass 357.1212373 Da and InChIKey SVQICVJPUYZYSA-UHFFFAOYSA-N. The calculated [M+H]+ expectation is m/z 358.1285138. Its molecular graph contains no specified stereocentre. These results identify an encoded structure, not an isolated chemical sample.

The book gives C19H21NO6 and approximately 359.4 g/mol for the named fisetin ether [11]. That pair is incompatible with retention of the specified fisetin scaffold during this single ether substitution. Two additional hydrogens would require another structural change. Ordinary protonation also does not resolve the discrepancy: the encoded amine-protonated form is C19H20NO6+, with InChIKey SVQICVJPUYZYSA-UHFFFAOYSA-O. The present graph therefore provides the explicit definition used throughout this communication.

Comparative Physicochemical Profile

Relative to fisetin, the derivative gains 71.123 g/mol, loses one hydrogen-bond donor and gains one acceptor. Calculated TPSA decreases by 7.76 Ų, from 111.13 to 103.37 Ų. Neutral-form Crippen cLogP increases by 0.2348, from 2.2824 to 2.5172. The change therefore introduces a basic functional group with a modest increase in calculated neutral lipophilicity; it does not demonstrate enhanced hydrophilicity or improved absorption.

The registered aminoethyl flavone has the same position-7 side chain but no free hydroxyl groups. It gives C19H19NO3, calculated TPSA 42.68 Ų, no donors and cLogP 3.4004. Its computed InChIKey, SMIVWJCBSIJRRL-UHFFFAOYSA-N, matches the NCATS record [9]. The three additional oxygens and hydroxyl pattern distinguish the candidate from that known graph, while explaining its more polar calculated profile.

Table 2. Calculated descriptors of three neutral encoded structures.

Property	Fisetin	Proposed ether	Registered aminoethyl flavone
Neutral formula	C15H10O6	C19H19NO6	C19H19NO3
MW, g/mol	286.239	357.362	309.365
Monoisotopic mass, Da	286.0477380	357.1212373	309.1364935
TPSA, Ų	111.13	103.37	42.68
Hydrogen-bond donors	4	3	0
Hydrogen-bond acceptors	6	7	4
Neutral Crippen cLogP	2.2824	2.5172	3.4004

Values were calculated using RDKit 2025.09.6. The comparator is PubChem CID 121180, linked from the NCATS identity record [9]. These descriptors are not experimental or pH-dependent pharmacokinetic measurements.

Table 2. Calculated descriptors of three neutral encoded structures.

Property	Fisetin	Proposed ether	Registered aminoethyl flavone
Neutral formula	C15H10O6	C19H19NO6	C19H19NO3
MW, g/mol	286.239	357.362	309.365
Monoisotopic mass, Da	286.0477380	357.1212373	309.1364935
TPSA, Ų	111.13	103.37	42.68
Hydrogen-bond donors	4	3	0
Hydrogen-bond acceptors	6	7	4
Neutral Crippen cLogP	2.2824	2.5172	3.4004

Values were calculated using RDKit 2025.09.6. The comparator is PubChem CID 121180, linked from the NCATS identity record [9]. These descriptors are not experimental or pH-dependent pharmacokinetic measurements.

The book's approximate TPSA of 100 Ų is consistent with a rounded description of the new calculation, whereas its logP of approximately 1.5 differs from the current method's cLogP [11]. No retained candidate-specific procedure establishes the basis of the earlier value. Agreement in one approximate descriptor does not validate unmeasured solubility, membrane transport or antibody affinity.

Public Database Findings

The neutral and protonated candidate InChIKeys returned PubChem HTTP 404 responses specifying that no matching CID was found. Neutral-SMILES same_connectivity and same_tautomer searches and the protonated-SMILES same_connectivity search likewise returned an identity-search no-hit response. The parent key returned fisetin CID 5281614, and the comparator record returned CID 121180 with the expected formula and key. The positive-control connectivity search also included CID 5281614 [18]. This supports a dated, query-specific no-match finding in PubChem Compound.

PubChem’s substructure query returned 47 CIDs, below the requested cap of 100. All 47 retrieved formulae differed from C19H19NO6 and none of their InChIKeys had the candidate connectivity block SVQICVJPUYZYSA. Larger or differently substituted graphs can satisfy a substructure query without representing the target molecule. CID 141343708 retained a fisetin 7-O-aminoethyl linkage but contained a substituted piperazine and fluoroquinolone terminus; CID 122182131 was a 7-O-diethylaminoethyl quercetin 3-O-methyl ether [10,26]. The formula-only query reached its 1000-record cap and was not treated as a complete exclusion search.

Both candidate keys returned ChEBI no-match responses; the fisetin key retrieved CHEBI:42567. The ChEBI formula search returned six entries with different keys. ChEMBL returned zero molecules for each candidate key and retrieved fisetin as CHEMBL31574. Its exact-formula filter returned 121 records, all with available Standard InChIKeys; none had the candidate connectivity block. These outcomes support limited public-record absence under the reported keys. They do not establish worldwide novelty or rule out unindexed individual structures [19,20].

Table 3. Completed public database queries and their interpretation.

Resource	Observed result on 12 October 2026	Interpretation
PubChem identity	No hits for the two keys and three SMILES identity queries; known controls retrieved	Candidate not retrieved under the specified identity queries
PubChem substructure	47 CIDs; no matching formula or candidate connectivity block	Related graphs were found and distinguished from exact identity
ChEBI	No match for either candidate key; parent retrieved; six formula entries differed	Key-specific absence in the public index
ChEMBL	Zero results for each candidate key; parent retrieved; 121 formula entries differed	Key-specific absence with a complete returned formula shortlist
PubMed and Europe PMC	Two and 131 records for the recorded broad queries	Literature-discovery counts; selected sources assessed, not exhaustive structure exclusions
Public patent texts	Four specified publication texts inspected	Neighbouring disclosures identified; genus and claim coverage unresolved
Access-limited services	CAS Common Chemistry API 401; selected Espacenet and PATENTSCOPE pages 403	Access failures supply no evidence of absence

Table 3. Completed public database queries and their interpretation.

Resource	Observed result on 12 October 2026	Interpretation
PubChem identity	No hits for the two keys and three SMILES identity queries; known controls retrieved	Candidate not retrieved under the specified identity queries
PubChem substructure	47 CIDs; no matching formula or candidate connectivity block	Related graphs were found and distinguished from exact identity
ChEBI	No match for either candidate key; parent retrieved; six formula entries differed	Key-specific absence in the public index
ChEMBL	Zero results for each candidate key; parent retrieved; 121 formula entries differed	Key-specific absence with a complete returned formula shortlist
PubMed and Europe PMC	Two and 131 records for the recorded broad queries	Literature-discovery counts; selected sources assessed, not exhaustive structure exclusions
Public patent texts	Four specified publication texts inspected	Neighbouring disclosures identified; genus and claim coverage unresolved
Access-limited services	CAS Common Chemistry API 401; selected Espacenet and PATENTSCOPE pages 403	Access failures supply no evidence of absence

Counts describe individual live API queries. They are not the number of included studies and were not used to construct a PRISMA flow diagram. Query parameters, response dates, record identifiers and screening files are supplied in the additional Search Evidence archive.

Relationship to Retrieved Prior Art

No individualised disclosure of the exact coded connectivity was identified in the retrieved public evidence. Completed PubChem identity and ChEBI/ChEMBL key queries strengthen that limited observation. They coexist with clear precedents for the scaffold and modification strategy. The original comparator establishes the 7-O-dimethylaminoethyl ether motif on a flavone [9]; the newly retrieved fisetin conjugate establishes a 7-O-aminoethyl linkage on a scaffold retaining the three free hydroxyl groups [10]. Differences at the terminal nitrogen distinguish the graphs, but do not make the general aminoethyl-fisetin strategy new.

The inspected 2012 paper describes eight aminoethyl flavone hydrochlorides derived from luteolin or its additional B-ring hydroxyl analogue. Its piperidine, pyrrolidine, morpholine and diethylamino termini, 5-OH substitution and absence of a free 3-OH distinguish the displayed final compounds from AIIP-FIS-7O-DMAE-001 [5]. The 2015 quercetin 3-O-methyl ether programme and the 2024 protected mF7 scaffold provide further nearby chemical contexts [6,8]. The 2023 fisetin derivative study establishes earlier experimental anti-inflammatory and ER-stress work; its biological findings are not results for the present candidate [21]. The exact structures in the partly accessible series remain a priority for further exclusion work.

The selected patent texts contained broad fisetin or polyphenol analogue wording and substituent definitions [22,23,24,25]. WO2013025498A1 also describes B-ring dimethylamino analogues such as CMS-117, which have different nitrogen attachment and ring substitution from the coded ether [25]. No exact named candidate was identified in the inspected text. A text-level comparison cannot exclude depiction in an unreviewed drawing or resolve whether a generic formula might encompass the candidate. Cyclodextrin ether wording in formulation contexts remains distinct from a covalent aminoethyl substituent on fisetin.

Table 4. Retrieved neighbouring disclosures and interpretation.

Record	Relevant subject	Relationship or limitation
Registered aminoethyl flavone [9,18]	7-O-dimethylaminoethyl motif	Candidate retains three additional hydroxyls
Chiruta et al. [2]	B-ring nitrogen substitutions	Different nitrogen attachment and ring pattern
Kim et al. [5]	Eight 7-O-aminoethyl flavone hydrochlorides	Inspected drawings show 5-OH, no free 3-OH and different amine termini
Lee et al. [6]; CID 122182131 [26]	Quercetin 3-O-methyl ether aminoethyl derivatives	Additional 5-OH and modified 3-OH; complete paper structure set not screened
PubChem CID 141343708 [10]	Fisetin 7-O-aminoethyl conjugate	Substituted piperazine–fluoroquinolone terminus rather than N,N-dimethylamine
mF7 programme [8]	Protected fisetin bearing a 7-O-butyltriphenylphosphonium group	Different chain, terminal cation and hydroxyl protection
Luo et al.; Lu et al. [3,4]	7-Aminoalkyl and nitrogen-containing flavonoids	Established strategy; complete drawings not exhaustively excluded
Correia da Silva et al. [21]	Experimental anti-inflammatory fisetin derivatives	Earlier pharmacological programme; candidate-specific efficacy untested
van Acker et al. [7]	3,7-Substituted catechol flavones	Older series identified; exact full-series exclusion incomplete
WO2008021210A2 [22]	Fisetin analogue and generic substituent wording	Exact depiction and specialist genus interpretation remain unresolved
Other selected patent texts [23,24,25]	Pyrone and polyphenol analogue contexts	No exact named candidate in inspected text; no claim-coverage conclusion

Table 4. Retrieved neighbouring disclosures and interpretation.

Record	Relevant subject	Relationship or limitation
Registered aminoethyl flavone [9,18]	7-O-dimethylaminoethyl motif	Candidate retains three additional hydroxyls
Chiruta et al. [2]	B-ring nitrogen substitutions	Different nitrogen attachment and ring pattern
Kim et al. [5]	Eight 7-O-aminoethyl flavone hydrochlorides	Inspected drawings show 5-OH, no free 3-OH and different amine termini
Lee et al. [6]; CID 122182131 [26]	Quercetin 3-O-methyl ether aminoethyl derivatives	Additional 5-OH and modified 3-OH; complete paper structure set not screened
PubChem CID 141343708 [10]	Fisetin 7-O-aminoethyl conjugate	Substituted piperazine–fluoroquinolone terminus rather than N,N-dimethylamine
mF7 programme [8]	Protected fisetin bearing a 7-O-butyltriphenylphosphonium group	Different chain, terminal cation and hydroxyl protection
Luo et al.; Lu et al. [3,4]	7-Aminoalkyl and nitrogen-containing flavonoids	Established strategy; complete drawings not exhaustively excluded
Correia da Silva et al. [21]	Experimental anti-inflammatory fisetin derivatives	Earlier pharmacological programme; candidate-specific efficacy untested
van Acker et al. [7]	3,7-Substituted catechol flavones	Older series identified; exact full-series exclusion incomplete
WO2008021210A2 [22]	Fisetin analogue and generic substituent wording	Exact depiction and specialist genus interpretation remain unresolved
Other selected patent texts [23,24,25]	Pyrone and polyphenol analogue contexts	No exact named candidate in inspected text; no claim-coverage conclusion

The supported classification is a chemically specified, phytochemical-derived computational candidate with provisional exact-structure novelty and documented no-match results in three public resources. It is not confirmation of a newly synthesised chemical entity, a new pharmacological mechanism or freedom to operate. The source book’s actual public-release history remains unresolved.

4. Discussion

What the Reconciliation Contributes

The contribution is a traceable conversion of an AI-assisted conceptual proposal into a fixed molecular graph. Correct elemental composition and attachment-site assignment are prerequisites for subsequent searching, synthesis and property interpretation. The identity discrepancy illustrates a practical limitation of accepting an AI-assisted chemical narrative without verifying its graph. Correcting it prevents comparison of the proposed ether with an inadvertently different saturation state or substituent pattern.

Exact identity also cannot be inferred from shared wording, formula or an apparently similar antibody illustration. Aminoethyl substitution at 3-O, 3′-O or 4′-O would yield different connectivity even when elemental composition is unchanged. The source's inconsistent figure numbering and immunoglobulin labels do not resolve those chemical distinctions. Explicit SMILES and InChI specifications make the structure under discussion auditable, while leaving experimental identification as a separate task [15].

Chemical Rationale and Unresolved Properties

The tertiary amine offers a testable route to pH-dependent ionisation and potential salt formation. At the same time, the retained hydroxyl groups preserve a relatively polar scaffold. Neutral cLogP cannot determine its charge distribution or partitioning under physiological conditions, and no candidate pKa or logD has been measured. Established physicochemical heuristics are useful for context rather than guarantees of oral exposure or development suitability [27].

The numerical comparison does not support a claim that aminoethyl substitution necessarily improves both aqueous solubility and passive membrane diffusion. Those properties depend on different processes and must be measured together with stability and usable exposure. The known comparator lacks the candidate's remaining hydroxyl groups, making it a chemical control rather than a substitute for candidate-specific pharmacology. Retention of the catechol and 3-OH also leaves chemical and metabolic questions that are not answered by a two-dimensional descriptor calculation.

An ether should not be labelled an ester prodrug merely because it contains an aminoethyl chain. No relevant bioconversion or release of an active parent has been demonstrated. Before a prodrug mechanism is proposed, the intact derivative and any metabolites would need to be identified and related to an independently measured biological response. This communication makes no prediction of a particular metabolic pathway or therapeutic advantage.

Experimental Questions Arising from the Specification

The first experimental requirement is a physical sample whose purity and position-7 attachment are analytically established. High-resolution mass spectrometry would test the elemental-composition expectation, but mass alone would not distinguish positional ethers. Complementary one- and two-dimensional NMR, with correlations supporting the aminoethyl OCH2 attachment to the assigned position-7 environment, would be needed. The present communication supplies no reaction conditions, yields or spectra as completed results.

The retained catechol and 3-OH groups require chemical stability and assay-compatibility assessment. Any biological evaluation should establish the exposure to intact compound and assess cell viability. Screening guidance supports orthogonal readouts and controls for chemical interference [28]. Direct experimental work with other flavonols has also shown that some viability assays can be misleading [29]. Structural alerts can guide triage but do not establish inactivity or nonspecific behaviour for this candidate [30]. No candidate-specific interference result has been generated.

If antibody modulation is pursued, the hypothesis should be narrowed to a structurally defined antibody preparation and a reproducible binding question. Direct binding and competition measurements would be required before proposing altered antigen recognition. Isotype labels alone cannot establish selectivity for pathogenic rather than protective antibodies. The source illustration provides no affinity estimate, validated binding pose or evidence of autoimmune-disease modification. Clinical claims would require a substantially later evidence base.

Novelty and AI Traceability

The extended public search makes the novelty assessment more reproducible: it specifies the exact input graphs, records genuine no-hit responses, verifies known controls and distinguishes 47 substructure records from the target. The three public resources have different coverage and some shared source content, so their agreement is not a statistical estimate of the probability of novelty. The defensible statement is that the specified candidate was not retrieved by the reported public identity/key queries on 12 October 2026. Worldwide structural novelty remains provisional because registry coverage, supplementary drawings, the partly accessible research series and patent genera are incomplete.

The next literature priority is full-structure assessment of the 2000, 2015, 2016, 2020 and 2023 series, including their supplementary files. Licensed exact and substructure searches should then test the neutral graph, standardised parent forms, relevant salts and positional isomers, with specialist examination of potentially relevant patent drawings and Markush formulae. These are further verification steps, not searches claimed as completed here. The project code remains useful for traceability regardless of the eventual novelty outcome.

The book may itself affect the disclosure history if its relevant passage has previously been released. The supplied copy contains placeholder bibliographic fields and does not establish a verified public release date. This article's contribution should therefore be assessed as chemical reconciliation and reproducible evaluation, rather than assumed first proposal or first public disclosure of the derivative.

The historical AI contribution is limited by missing model and interaction records. No new model architecture, predictive training, optimisation benchmark or AI advantage is demonstrated. The reproducible component is the deterministic chemical analysis. A prospective AI-design study would preserve prompts, model access details, all proposed candidates, rejection criteria and human corrections, then compare selections against conventional design baselines and measured outcomes. That would constitute a different study from the present audit [12,13].

5. Conclusions

AIIP-FIS-7O-DMAE-001 is the project code for fisetin 7-O-[2-(dimethylamino)ethyl] ether, an encoded computational candidate with neutral formula C19H19NO6 and molecular weight 357.36 g/mol. Its chemical specification corrects the source formula and mass mismatch. The reported calculations describe a modest increase in neutral lipophilicity and a reduction in polar surface area without establishing improved exposure. The completed PubChem, ChEBI and ChEMBL queries did not retrieve the exact candidate, whereas substructure and literature searches identified substantial neighbouring modification precedent. These results support provisional exact-structure novelty within the stated public search scope. Synthesis, chemical characterisation and experimental pharmacology remain necessary; the code and database outcomes establish no antibody binding, efficacy or clinical utility.