From Lost Smell to a Testable Research Program

NR-100 Hydrazine-Free Discovery Platform Provisional Primary Hypothesis NR-101 and Matched Analogs NR-102 Through NR-106

Research Status and Safety Statement

No experimental data currently establish that any NR-100-series compound:

• Binds to, activates, or inhibits human 5-HT2A or another molecular target.

• Possesses low intrinsic efficacy, reduced Gi signaling, favorable bias, or useful receptor selectivity.

• Promotes structural or functional neuroplasticity.

• Regenerates olfactory sensory neurons or restores smell or taste.

• Avoids psychedelic, dissociative, sedative, cardiovascular, metabolic, genotoxic, or local-tissue effects.

• Crosses the blood-brain barrier or reaches a useful intracellular receptor compartment.

• Is suitable for intranasal, oral, systemic, repeated, or human administration.

• Is patentable, free to operate, manufacturable, stable, or acceptable to a regulatory authority.

The structural records in this document are provisional. The controlled two-dimensional molecular drawing and independently verified SDF or MOL record must become authoritative before synthesis or vendor transfer. Text identifiers must be generated from that controlled structure by at least two independent chemical-informatics systems.

No synthesis instructions, dosing instructions, or formulation instructions are provided. Any future laboratory work must occur through qualified organizations under applicable chemical-safety, controlled-substance, animal-welfare, human-subjects, manufacturing, and regulatory requirements.

Author's Perspective

Persistent loss or distortion of smell and taste can substantially affect nutrition, environmental safety, emotional well-being, memory, social interaction, and quality of life.

My interest in olfactory research arose from my personal experience with prolonged smell and taste loss following COVID-19. After living with these impairments for approximately two years, I experienced substantial recovery in 2022. That experience led me to investigate biological pathways involved in neural plasticity, sensory-neuron renewal, inflammatory injury, and olfactory repair.

NeuroRasool emerged from this investigation as a proposed molecular research program. Version 4.3 retains the NR-100 platform and Version 4.2 scientific controls, then adds formal interface governance for parallel structure-control and analytical-readiness workstreams. It preserves bounded models, functional olfactory criteria, curve and recovery rules, and the separation between exploratory scores and safety gates. Structural refinement and modeled values are not proof of safety.

Abstract

Persistent post-viral olfactory dysfunction is biologically heterogeneous and may involve sustained immune activation, loss or dysfunction of olfactory sensory neurons, altered support-cell biology, impaired progenitor differentiation, disrupted epithelial architecture, abnormal axonal targeting, and central-network changes. A general cortical psychoplastogen cannot be assumed to correct these processes. Any NeuroRasool candidate must therefore demonstrate direct, reproducible benefit in qualified human olfactory systems before an olfactory-treatment claim is considered.

Version 4.3 retains the hydrazine-free NR-100 discovery platform and makes no molecule substitution. NR-101 remains the provisional first test article. Its three-carbon tether terminates at the morpholine nitrogen, while matched analogs test basicity, polarity, aromatic steric demand, and linker geometry. Hydrazine removal and reduced aromatic bulk eliminate or reduce specific design concerns, but they do not establish lower integrated toxicity, hERG risk, phospholipidosis, tissue accumulation, target engagement, or regulatory advantage.

The updated pharmacology strategy rejects a simple Gq-therapeutic versus beta-arrestin-hallucinogenic model. Recent evidence supports evaluating absolute 5-HT2A efficacy, Gq and G11 recruitment, non-canonical Gi-family signaling, beta-arrestin recruitment, receptor internalization, kinetics, receptor reserve, intracellular access, 5-HT1A and 5-HT1B contributions, and active metabolites. Low efficacy and reduced Gi activity are prioritized as testable design goals, but neither is treated as a universal rule.

The immediate objective is to determine whether an NR-100-series compound has reproducible target-relevant pharmacology and direct olfactory-model activity that can be separated from 5-HT2B agonism, broad aminergic activity, hERG inhibition, reactive metabolism, genotoxicity, phospholipidosis, mitochondrial injury, psychoactive liability, and nasal-tissue toxicity. Version 4.3 retains the quantitative and safety tools from Version 4.2 and adds a controlled feedback loop between molecular-property review and analytical RFP design. These tools organize decisions; they do not replace measured data.

Keywords: NeuroRasool, NR-101, hydrazine-free analog, 5-HT2A, psychoplastogen, low-efficacy agonism, Gi signaling, 5-HT1B, olfactory dysfunction, post-viral anosmia, human olfactory organoid, medicinal chemistry

Version 4.3 Executive Decisions

1. Scope and Program Reset

Version 4.3 is a workstream-governance update to Version 4.2. It retains NR-101 through NR-106 and all scientific evidence requirements, while adding explicit synchronization, dependency control, vendor-bidding boundaries, and joint release criteria for structure control and analytical readiness.

1.1 What is retained

• The olfactory-recovery mission and the author's emphasis on persistent post-viral dysfunction.

• A non-indole naphthalene research identity distinct from psilocin's indole scaffold.

• Measured-exposure requirements before accepting potency or inactivity.

• Multidimensional receptor pharmacology and strict 5-HT2B risk management.

• Hydrazine, reactive-metabolite, hERG, CYP, lysosomal-trapping, phospholipidosis, mitochondrial, and nasal-tissue safety gates.

• Qualified human olfactory models, donor-level inference, biospecimen governance, raw-data access, and version-controlled decisions.

• A program-level scaffold pause when qualified evidence does not support continued investment.

1.2 What is replaced

• NR-001 is no longer an active lead. It remains only a legacy and hazard comparator.

• NR-001P is frozen indefinitely. A phosphate derivative will not be pursued unless a validated active parent has a specific delivery limitation that a cleavable polar precursor can plausibly solve.

• The tert-butyl naphthalene family NR-002 through NR-006 is moved to the historical comparator set.

• The active program becomes NR-101 through NR-106, with smaller or electronically different aromatic substituents and hydrazine-free terminal amines.

• The optimization objective changes from presumed Gq bias to measured low or moderate intrinsic efficacy, reduced undesirable signaling, direct plasticity, and integrated safety.

1.3 Immediate research question

1.4 Version 4.3 evidence-to-decision controls

• NR-101 remains the primary provisional molecule; no structure is promoted or replaced by this update.

• Track A structure and M0 property records and Track B analytical requirements are version-linked and reviewed through a continuous change-control loop.

• A material property change automatically reopens every affected recovery, stability, calibration, matrix, and mass-balance requirement before vendor scope lock.

• Vendor market sounding and feasibility questions may begin from controlled assumptions, but no binding performance guarantee or synthesis release may rely on an unfrozen structure record.

• Structural refinement is described as removal or reduction of specific concerns, never as demonstrated safety.

• The morpholine nitrogen is treated as a testable aminergic-anchoring hypothesis rather than a necessary pharmacophore.

• Reduced 5-HT2A Gi recruitment remains an experimental counter-screen and is not accepted as proof of non-hallucinogenic action in humans.

• Calculated pKa, logD, and trapping risk belong to M0; experimental pKa, logD, disposition, and trapping require qualified material at M1 or later.

• UGT- and SULT-mediated phenolic conjugation, hepatocyte clearance, and metabolite identification are explicit early ADME requirements.

• Illustrative ionization and lysosomal-partition equations are included with explicit assumption limits; untraceable candidate predictions are excluded.

• Human olfactory advancement requires a composite of viable lineage, architecture, barrier or support-cell integrity, and functional response rather than one marker.

• 5-HT2B and hERG findings are interpreted against projected unbound exposure and active metabolites, not by isolated nominal-concentration cutoffs.

• Weighted portfolio scoring is subordinate to hard identity, exposure, genotoxicity, cardiac, and tissue-safety gates.

2. Unmet Need and Biological Context

2.1 Persistent post-viral olfactory dysfunction

Persistent smell loss and distortion after viral illness can reflect several biological states rather than one uniform disease. Human post-COVID biopsy work has associated persistent dysfunction with immune-cell infiltration, altered epithelial gene expression, and reduced olfactory sensory-neuron representation even after detectable viral material has cleared [10]. Injured human olfactory epithelium may also show altered progenitor behavior, support-cell dysfunction, respiratory metaplasia, and loss of organized neuronal identity [11, 12].

A credible program must separate patients and models by phenotype. Normosmia, persistent quantitative loss, parosmia, phantosmia, inflammatory sinonasal disease, recovered post-viral state, and nonviral injury should not be pooled into one biological category.

2.2 Regeneration is not one endpoint

Olfactory recovery may require coordinated improvement across basal-cell activation, lineage commitment, neuronal maturation, cilia formation, odorant-transduction machinery, epithelial barrier function, axonal organization, inflammation, and central processing. Increasing one marker or producing nonspecific neurite extension is insufficient.

2.3 Why serotonergic plasticity remains a hypothesis

Serotonergic psychedelics and several non-hallucinogenic analogs can promote structural and functional plasticity in experimental neuronal systems [1, 5, 6]. This supports exploration of related mechanisms, but no published study establishes that a 5-HT2A psychoplastogen regenerates human olfactory epithelium or treats persistent post-viral smell loss. NeuroRasool must prove its indication in the intended tissue context.

3. Evidence Update Through July 2026

3.1 Low efficacy may distinguish some non-psychedelic agonists

A 2025 comparison of hallucinogenic and non-hallucinogenic 5-HT2A ligands reported that lisuride, tabernanthalog, and IHCH-7079 were differentiated more consistently by low signaling efficacy than by a single calculated bias factor [4]. The authors noted that a larger ligand panel is needed. No universal threshold is established because efficacy depends on receptor density, assay amplification, transducer complement, reference agonist, and analysis model. The result supports measuring an efficacy ceiling, not claiming that low efficacy predicts the human subjective profile.

3.2 Gi signaling is a new but not yet universal discriminator

A 2026 structural and pharmacological study reported that non-canonical 5-HT2A-Gi signaling was necessary for hallucinogenic effects in the experimental systems examined and described a Gq-favoring ligand with reduced hallucinogenic-like behavior in mice [3]. This is a major advance and should be included directly in the NR-100 assay panel. It should not be treated as proof that every Gi-sparing ligand is non-hallucinogenic in humans.

3.3 Zalsupindole strengthens the development precedent

Zalsupindole, also known as DLX-001, is an isotryptamine-derived neuroplastogen reported to produce structural and behavioral effects in preclinical systems without hallucinogenic or dissociative effects [5]. Public clinical-development reports make it a useful program precedent, but public development status and early reports are not evidence that the NeuroRasool scaffold, mechanism, indication, or safety profile will translate.

3.4 Polypharmacology may contribute to persistent effects

Mouse studies published in 2025 linked 5-HT1B signaling to psilocybin-related cortical communication and selected persistent behavioral effects while distinguishing those findings from the acute head-twitch response [8, 31]. A separate investigation of 5-halogenated DMT analogs found that 5-Br-DMT produced little or no head twitch, promoted dendritic growth, and favored 5-HT1A over 5-HT2A, although it also caused hypothermia and reduced locomotion [7]. These findings support measuring 5-HT1A and 5-HT1B rather than treating them as irrelevant off-targets.

3.5 Intracellular 5-HT2A access remains important

Work published in 2023 implicated intracellular 5-HT2A receptors in psychedelic-associated plasticity in the systems tested [1]. Therefore, total uptake, neutral fraction, free intracellular exposure, lysosomal sequestration, and surface-restricted versus intracellular signaling remain essential. A polar compound can have low access; a lipophilic basic compound can show high total uptake while remaining trapped and unavailable to the intended target.

4. NR-100 Molecular Design and Physicochemical Model

4.1 Design objective

The NR-100 series is intended to test whether a hydrazine-free, non-indole naphthalene pharmacophore can produce useful serotonergic or alternative pharmacology with a safer integrated profile. The series is deliberately small. It changes one or two major variables at a time so that activity, inactivity, and liability can be interpreted.

4.2 Primary structural reset

4.3 Why morpholine leads the first comparison

NR-101 contains a three-carbon tether that terminates at the morpholine nitrogen. Morpholine provides an ionizable tertiary amine and may offer lower basicity and greater polarity than matched pyrrolidine, piperidine, or dimethylamine analogs. This may change neutral fraction, membrane passage, aminergic target engagement, off-target binding, hERG risk, lysosomal sequestration, and phospholipidosis. It does not prove improvement in any of those properties. Experimental pKa, logD, permeability, intracellular free exposure, trapping, and pharmacology are required.

4.4 Why isopropyl replaces tert-butyl

The tert-butyl group was previously used as a steric design element, but it may impose unnecessary lipophilicity, aromatic bulk, metabolic persistence, nonspecific binding, and tissue retention. Isopropyl is a smaller matched steric probe. NR-104 and NR-105 further test whether cyclopropyl or fluoro substitution offers a better balance. The program must not interpret smaller substitution as automatic safety or target selectivity.

4.5 No claim of novelty

A keyword and exact-text web search did not identify an obvious public record for the provisional NR-101 string, but that is not a structure search and does not establish novelty, patentability, or freedom to operate. A professional substructure, similarity, Markush, sequence, salt, stereochemistry, and use-claim search is required before public disclosure or investment claims.

4.6 Quantitative ionization and trapping model

For a monoprotic base at equilibrium, the neutral fraction may be estimated from pKa and pH. The calculation is useful for comparing hypotheses across nasal, cytosolic, and physiological conditions, but it is not a permeability measurement.

An idealized pH-partition ratio can also frame lysosomal-trapping risk by comparing an acidic lysosomal compartment with cytosol. It assumes freely diffusible, nonsaturating, monoprotic behavior and excludes binding, transport, metabolism, organelle volume, membrane potential, and kinetic limitations.

5. Provisional Chemical Identities

Formula and exact-mass values are provisional calculations from the written structures and must be independently confirmed. The position numbering implied by each text string must be reconciled with the controlled 2D drawing before synthesis.

5.1 NR-101: provisional primary test article

NR-101 is not a nominated clinical candidate. It is the first compound selected for controlled testing because its design changes two high-priority medicinal-chemistry variables from the earlier program: hydrazine removal and reduction of aromatic steric bulk. The propyl tether terminates at the morpholine nitrogen. That ionizable nitrogen tests a possible aminergic anchoring interaction; it is not established as necessary or sufficient for target engagement.

Design advantages are limited to removal of the N-N motif, lower predicted basicity than common strong amines, and lower predicted hydrophobic burden than the tert-butyl predecessor. These are not safety findings. Potential liabilities include weak target anchoring, rapid UGT or SULT phenolic conjugation, broad aminergic activity, oxidation, hERG inhibition, lysosomal accumulation, phospholipidosis, insufficient free intracellular exposure, and complete absence of useful activity.

5.2 Matched controls

NR-102 tests whether a strongly basic but polar cyclic amine improves target engagement without excessive lipophilic trapping. NR-103 tests a smaller acyclic strong base. NR-104 isolates aromatic steric shape. NR-105 tests a lower-burden electronic substituent. NR-106 tests whether a two-carbon linker better fits an aminergic binding geometry. The first campaign should not expand beyond these compounds unless Gate 0B establishes actionable pharmacology.

5.3 Controlled M0 property record

Version 4.3 does not promote the illustrative pKa or logD values in the submitted draft to controlled candidate properties. The M0 record must be regenerated from the authoritative structures under a documented model workflow.

6. Target Product Profile

The target product profile is a discovery objective, not a claim that the molecule can meet it. It is intentionally route-neutral and mechanism-flexible.

7. Central Hypotheses

Failure of any individual hypothesis must be documented. Failure of H1 under qualified conditions triggers a scaffold-level pause. Failure of H5 prevents an olfactory indication even if another pharmacology program remains scientifically possible.

8. Structural Pharmacology and Mechanistic Model

8.1 Receptor-state ensemble

Docking and molecular simulation should use a receptor-state ensemble rather than one favored structure. The minimum set should include active Gq-associated 5-HT2A, available Gi-associated structures, inactive 5-HT2A, and relevant 5-HT2B and 5-HT2C structures or validated models. Alternative protonation states, structured waters, membrane composition, and receptor flexibility must be considered [2, 3].

8.2 Conserved aminergic anchoring

A protonated amine may interact with Asp1553.32 in 5-HT2A and related aminergic receptors. The NR-100 series retains one ionizable nitrogen to test this possibility, but the interaction and its necessity are unproven. Morpholine may provide a smaller protonated fraction than stronger amines. The actual microscopic pKa, neutral fraction, binding mode, residence time, and functional consequence must be measured or modeled with uncertainty.

8.3 Efficacy is context-dependent

A ligand does not possess one universal Emax. Apparent efficacy changes with receptor expression, transducer availability, assay amplification, coupling efficiency, incubation time, reference agonist, and data model. NR-100 compounds must therefore be compared in direct recruitment assays and orthogonal downstream assays using the same reference panel.

8.4 Current mechanistic model

9. Receptor and Signaling Strategy

9.1 Primary receptor panel

• Human 5-HT2A binding or validated target engagement.

• Human 5-HT2A agonist and antagonist modes.

• Direct Gq and G11 recruitment.

• Gi1, Gi2, Gi3, GoA, and GoB recruitment where technically feasible.

• IP1 accumulation and calcium mobilization as amplified orthogonal assays.

• Beta-arrestin-1 and beta-arrestin-2 recruitment.

• Internalization, desensitization, onset, duration, washout, and residence time.

• Human 5-HT2B and 5-HT2C functional assays.

• 5-HT1A and 5-HT1B functional pharmacology.

9.2 Expanded counter-screen

The first expanded panel should include 5-HT1D, 5-HT6, 5-HT7, dopamine D1 and D2 families, alpha1 and alpha2 adrenergic receptors, histamine H1, monoamine transporters, sigma receptors if predicted, hERG, and any high-confidence computational alerts. A lipophilic protonatable amine must be presumed capable of broad aminergic pharmacology until tested.

9.3 Provisional efficacy interpretation

No universal Emax threshold is established. For internal triage, the program may define low, moderate, and high efficacy bands relative to the same reference agonist in direct recruitment assays. These bands are assay-specific and may not be transferred between vendors or platforms without bridging.

These ranges are internal design bands, not clinical or regulatory standards. Low Emax values should be reported with confidence intervals and limits of quantification rather than a coefficient of variation that becomes unstable near zero.

10. Physicochemical, ADME, and Exposure Strategy

10.1 Calculated properties before synthesis

• Microscopic and macroscopic pKa with uncertainty.

• Neutral fraction at pH 5.5, 6.0, 6.5, 7.0, and 7.4.

• logP, logD, topological polar surface area, solubility, and intrinsic solubility.

• Permeability, efflux probability, brain-partition probability, and unbound fraction.

• Conformational ensemble, ligand strain, molecular volume, and aromatic planarity.

• pKa-informed lysosomal-trapping estimates and complementary phospholipidosis, hERG, CYP, reactive-metabolite, and pan-assay-interference alerts [29].

• Predicted sites of phenolic glucuronidation and sulfation, reported as risk hypotheses rather than measured clearance [30].

10.2 Measured properties after synthesis

• Thermodynamic and kinetic solubility across relevant pH and media.

• Experimental pKa and logD.

• Plasma stability, microsomal stability, hepatocyte clearance, and metabolite identification.

• UGT-supported glucuronidation and SULT-supported sulfation assessment with appropriate cofactors and controls.

• Formation, stability, and recovery of major phenolic conjugates with parent-plus-metabolite mass balance.

• Plasma protein binding and blood-to-plasma partitioning.

• Passive and cellular permeability, transporter effects, and recoverable free intracellular exposure.

• Total cellular accumulation separated from lysosomal sequestration using an orthogonal trapping assessment.

• Chemical, photochemical, freeze-thaw, autosampler, and assay-incubation stability.

10.3 Exposure is part of every biological result

Nominal concentration is insufficient. Parent compound and important metabolites should be measured at time zero and assay endpoint in the actual buffer or medium. Where cells are used, extracellular, cell-associated, plate-rinse, precipitate, and major metabolite fractions should support a fit-for-purpose mass balance. Inactivity cannot be accepted when the compound is absent, precipitated, adsorbed, degraded, or trapped outside the relevant target compartment.

Each term must be operationally defined. Cell-associated mass is not automatically target-accessible mass, and metabolite terms should be converted to a stated parent-equivalent basis when used in a recovery calculation.

10.4 Phenolic disposition and trapping decision rule

The retained naphthol makes glucuronidation and sulfation plausible early-clearance routes, but evidence from model naphthols does not establish the rate, responsible isoforms, or consequence for NR-101 [30]. Qualified material must be tested for parent disappearance, conjugate formation, hepatocyte clearance, intracellular free exposure, and recovery. Isoform attribution should be made only from appropriate recombinant, inhibitor, or reaction-phenotyping evidence. Rapid conjugation is not automatically disqualifying if adequate active exposure can be demonstrated, while high total cellular uptake is not favorable when it is explained primarily by lysosomal sequestration [29].

11. Delivery and Formulation Position

11.1 Route-neutral lead discovery

Version 4.3 does not assume intranasal delivery. The initial question is whether an active and sufficiently safe parent compound exists. Route selection follows pharmacology, exposure, tissue-distribution, local-tolerance, and indication data.

11.2 Intranasal development conditions

Intranasal work becomes justified only if the parent compound has useful pharmacology, acceptable early safety, a credible reason for local nasal or nose-to-brain exposure, and an exposure limitation that formulation can address. Human nasal deposition, mucociliary clearance, epithelial permeability, local metabolism, and tissue injury must then be evaluated explicitly.

11.3 Prodrug position

No phosphate or other prodrug is part of the initial NR-100 campaign. A prodrug requires a validated parent, a named delivery problem, evidence that the promoiety addresses that problem, conversion data in the intended tissue, and proof that the released parent reaches the relevant compartment. Increased aqueous solubility alone is not sufficient.

12. Stage-Gated Development Architecture

A later-stage result cannot repair an invalid earlier gate. For example, an apparent organoid effect does not override an identity error, poor recovery, cytotoxic concentration, or 5-HT2B hazard.

13. Gate 0A: Structure, Identity, and Analytical Recovery

13.1 Structure-transfer requirements

• Medicinal-chemist approval of the authoritative two-dimensional structure.

• SDF and MOL records generated from the authoritative structure.

• Canonical and isomeric SMILES, InChI, InChIKey, formula, exact mass, charge, tautomer, salt, and stereochemical status generated independently by two systems.

• Hash-controlled files and a structure-verification report.

• Vendor confirmation that registration, synthesis, analytical records, plates, and reports use the approved code and structure.

13.2 Synthesis-release requirements

Synthetic feasibility must be reviewed without assuming a route. The vendor package must address impurity and regioisomer control, residuals, phenol and amine stability, salt form, analytical separation, scale, and reserve material. Biological release requires confirmed identity, fit-for-purpose purity, and stability.

13.3 Analytical method qualification

The analytical lower limit of quantification should ordinarily be no greater than 20% of the lowest decision-relevant concentration. Separate ranges may support pharmacology and mass-balance work. Performance must be demonstrated in each relevant matrix.

13.4 Parallel-Track Interface Control

Track A and Track B may proceed concurrently to shorten calendar time, but they are not independent workstreams. Track A controls the authoritative molecular structures, identifiers, file hashes, and M0 property hypotheses. Track B converts those controlled records into analytical requirements, matrix-recovery standards, mass-balance expectations, reference-ligand rules, and vendor deliverables. The operating objective is traceable alignment between the tracks, not an unsupported claim of perfect calibration.

13.5 Continuous Feedback Loop and Freeze Points

1. Track A publishes each candidate record as a controlled package containing the authoritative structure, parser agreement, identifiers, hashes, model inputs, outputs, uncertainty, software version, reviewer, and effective date.

2. Track B maintains a dependency matrix that links every compound-specific analytical requirement to the exact Track A record and states whether the input is provisional, conditionally accepted, or frozen.

3. A material Track A change automatically triggers a Track B impact assessment before further scope lock, method qualification, vendor guarantee, or synthesis decision. The assessment must identify affected matrices, concentration ranges, controls, recovery expectations, stability conditions, and mass-balance terms.

4. A material Track B feasibility finding feeds back to Track A. Poor recovery, adsorption, precipitation, instability, ion suppression, or incompatible reference-ligand behavior must be evaluated as a potential property or structure issue rather than hidden by a method exception.

5. Vendor market sounding, capability review, reference-ligand feasibility questions, and indicative pricing may begin before design freeze when assumptions are explicit. Final scope, fixed performance commitments, and synthesis release require frozen inputs.

6. The program approver closes 0A-S and 0A-R only when both tracks reference the same candidate codes, controlled structures, hashes, property-record versions, analytical requirements, deviations, and unresolved-risk register.

14. Gate 0B: Scaffold Viability

Gate 0B asks whether the NR-100 series has reproducible target-relevant activity under qualified exposure conditions. It occurs before advanced signaling, organoids, formulation, animal efficacy, or broad chemistry expansion.

14.1 Minimum viability panel

• Human 5-HT2A binding or validated target engagement.

• Human 5-HT2A agonist and antagonist functional modes.

• Human 5-HT2B and 5-HT2C functional activity.

• At least one orthogonal 5-HT2A functional assay for decision-driving findings.

• 5-HT1A and 5-HT1B functional screening.

• Early aminergic counter-screen and hERG triage.

• Cytotoxicity in the same assay environment.

• Measured parent exposure at time zero and endpoint.

• Solubility, precipitation, adsorption, aggregation, and interference controls.

14.2 Scaffold-level decisions

15. Multidimensional Pharmacology

15.1 Assay architecture

Direct transducer-recruitment assays should serve as the primary platform for pathway efficacy and bias characterization. IP1 and calcium may provide orthogonal downstream confirmation but should not independently establish bias. Each candidate should be studied with at least eight concentrations, duplicate technical wells, and at least three independent experiments unless a documented feasibility decision specifies a different design.

15.2 Reference set

15.3 Bias calculations

Bias factors should be calculated only when the data support the operational model and common reference assumptions. Confidence intervals must be reported. A compound should not be called biased when efficacy is too low for stable estimation, reference fits fail, receptor reserve differs materially, or separate platforms cannot be bridged.

15.4 Kinetics and location

Endpoint assays can conceal transient activation, delayed recruitment, desensitization, and time-dependent bias. Priority compounds require time-course and washout studies. Surface-restricted receptor systems, membrane-impermeant controls, or compartment-sensitive biosensors should be used where feasible to distinguish surface and intracellular activation.

16. Neuroplasticity Testing

16.1 Plasticity must be measured directly

• Dendritic complexity and neurite extension with automated morphology and blinded quality control.

• Spine density or synaptic puncta where the system permits.

• Functional synaptic activity, network connectivity, or electrophysiology.

• Temporal persistence after washout.

• Cell viability, stress responses, and proliferation measured in parallel.

• Mechanism controls using receptor antagonists, knockdown or knockout, and pathway-selective interventions where valid.

16.2 Cortical assays are supportive, not indication-defining

Cortical neuronal systems can determine whether a compound has general psychoplastogenic activity and can help compare with published ligands. They cannot establish repair of olfactory epithelium. A candidate that is active cortically but inactive or harmful in olfactory models cannot advance for the olfactory indication.

16.3 Avoiding false plasticity

Apparent neurite growth may reflect stress, altered cell density, selective survival, segmentation artifacts, or proliferating non-neuronal cells. Experiments must include concentration-matched viability, cell-count normalization, morphology quality control, predefined image-analysis rules, and raw-image retention.

17. Human Olfactory Translation

17.1 Model hierarchy

1. Qualified human olfactory epithelial cultures for early toxicity and lineage effects.

2. Donor-derived air-liquid-interface or organoid models with documented olfactory identity.

3. Injury or inflammatory models with a measurable rescue window.

4. Phenotype-linked persistent post-viral donor models for confirmatory translation.

5. Functional odorant-evoked or electrophysiological testing where feasible.

17.2 Required identity domains

17.3 Functional-rescue composite

A qualified olfactory model should predefine an injury window, positive control, negative control, maturity criteria, and a panel of validated odorants or physiological stimuli appropriate to the model. No single odorant, dye, marker, or culture recipe is mandated across platforms. Calcium imaging, electrophysiology, cAMP-linked transduction, or another validated physiological assay may establish function when paired with mature olfactory identity and concentration-matched viability.

The index is an exploratory normalized contrast, not a universal acceptance threshold. It becomes unstable when the healthy-control and injury means are close, so the raw effect, denominator uncertainty, donor-level interval, and assay dynamic range must also be reported.

17.4 Mechanism-neutral advancement

An NR-100 candidate must not advance solely because it binds 5-HT2A, has reduced Gi recruitment, or increases cortical neurites. Advancement requires a reproducible composite benefit in qualified olfactory models, such as preserved barrier function, viable progenitor activation, appropriate neuronal differentiation, mature sensory-neuron and ciliary identity, reduced pathological inflammatory signaling, organized architecture, and functional odorant response. If the benefit is not mediated by the proposed receptor mechanism, the program must reclassify the pharmacology rather than discard a reproducible tissue benefit or force a 5-HT2A narrative.

17.5 Donor-level inference

The donor is the primary biological replicate for advancement-critical inference. Wells, fields, organoids, repeated measures, and batches from one donor are not independent donors. Primary analyses must retain donor identity and model donor effects. Pooling may be used only for separately justified technical development and cannot replace donor-level results.

17.6 Biospecimen governance

Every tissue stream requires written institutional or IRB determination, consent or authorization as applicable, coding and privacy controls, chain of custody, phenotype metadata, collection-site documentation, ischemic or excision-to-medium timing, surgical exposure records, processing deviations, and reserve capacity. Coded tissue should not be presumed outside human-subjects requirements without institutional determination.

18. Safety and Toxicology

18.1 Structural improvement is not proof of safety

Removing hydrazine eliminates one important alert but introduces no presumption of general safety. Reducing aromatic bulk does not demonstrate lower hERG risk, phospholipidosis, nonspecific binding, or tissue accumulation. NR-100 compounds remain aromatic phenols with protonatable amines. Their risks may include broad aminergic pharmacology, hERG inhibition, CYP interactions, N-dealkylation, amine oxidation, UGT- or SULT-mediated phenolic conjugation, reactive metabolites, lysosomal trapping, phospholipidosis, mitochondrial effects, tissue accumulation, and local nasal injury.

18.2 Safety cascade

18.3 5-HT2B risk rule

Any reproducible 5-HT2B agonism above assay noise triggers an exposure-based analysis. Binding affinity must not be divided by functional potency as if the quantities were interchangeable. Before PK exists, same-mode functional selectivity over 5-HT2A is only a temporary medicinal-chemistry proxy. Once exposure data exist, the primary interpretation compares the lowest relevant 5-HT2B functional effect concentration with projected unbound human exposure, incorporating efficacy, duration, accumulation, uncertainty, and active metabolites [16].

A larger margin is preferred, and an aspirational 30-fold discovery margin may be used for prioritization. No isolated ratio, including 30-fold, establishes cardiovascular safety. Final criteria require toxicology, pharmacokinetic, cardiology, and regulatory review.

18.4 Integrated stop conditions

• Reproducible DNA-reactive or mammalian genotoxic mechanism that cannot be separated by redesign.

• Strong 5-HT2B agonism at relevant exposure.

• hERG or broader cardiac risk without a credible exposure margin.

• Activity only at cytotoxic, precipitating, or assay-interfering concentrations.

• Substantial phospholipidosis or lysosomal trapping that explains apparent uptake or phenotype.

• Severe mitochondrial, hepatocyte, or nasal epithelial injury.

• Active metabolite with unacceptable pharmacology or toxicity.

19. Behavioral and Psychoactive-Liability Assessment

19.1 No single animal assay proves non-hallucinogenic action

The mouse head-twitch response is useful for comparing many 5-HT2A agonists but is not a human subjective-experience assay. Lack of head twitch can result from low exposure, species pharmacology, competing receptor activity, sedation, hypolocomotion, hypothermia, motor impairment, or antagonist-like behavior. A negative result requires PK and a broader behavioral interpretation.

19.2 Recommended integrated battery

• Head-twitch response with blinded or automated scoring and receptor-antagonist controls.

• Locomotion, coordination, startle or sensorimotor gating, temperature, autonomic measures, and observational battery.

• Drug discrimination against an appropriate psychedelic comparator when justified.

• EEG or other translational pharmacodynamic measures where validated.

• PK and unbound brain exposure aligned with behavioral time points.

• Assessment of sedation, anxiogenesis, agitation, seizure risk, and cognitive disruption.

• Abuse-liability and dependence planning if pharmacology or scheduling analysis indicates concern.

The dose and exposure range should be justified from pharmacology, tolerability, target engagement, and the maximum feasible or tolerated exposure. A fixed 100-fold multiple or a fixed comparator dose is not a universal requirement and may be scientifically or ethically inappropriate.

19.3 Human conclusion

Only appropriately designed human studies can determine whether a compound produces hallucinatory, perceptual, dissociative, or impairing subjective effects in humans. Preclinical data may support a reduced-risk hypothesis but cannot establish a human non-hallucinogenic claim.

20. Candidate Ranking and Portfolio Governance

20.1 Weighted evidence model

Potency alone must not determine the lead. A moderately potent compound with low efficacy, clean 5-HT2B pharmacology, free intracellular exposure, direct olfactory benefit, and a usable safety window may outrank a potent but unsafe or behaviorally active compound.

Weights are a starting governance framework and must be locked before reviewing the decision-driving dataset. Changes require written justification and sensitivity analysis. Identity failure, uninterpretable exposure, material genotoxicity, unacceptable 5-HT2B or cardiac risk, and severe target-tissue toxicity are hard gates that cannot be averaged away by a high total score.

20.2 Portfolio classifications

21. Data Quality, Statistics, and Reproducibility

21.1 Minimum pharmacology standards

• At least eight concentrations spanning both plateaus when technically possible.

• Duplicate technical wells and at least three independent experiments.

• Reference agonist, vehicle, inactive or antagonist control, and cytotoxicity control.

• Z-prime of at least 0.5 when applicable, with documented signal window and control performance.

• Full concentration-response curves, raw plate data, exclusions, model equations, residuals, and confidence intervals.

• Measured compound exposure and run-specific recovery for decision-driving assays.

• No averaging of failed runs into apparently acceptable performance.

21.2 Curve modeling and assay quality

Concentration-response work should ordinarily use a prespecified four-parameter logistic model when the biology and data support both plateaus. Alternative models require justification. Parameter estimates, confidence intervals, residuals, convergence diagnostics, and any constrained parameters must be reported.

For high-throughput screens, Z-prime may summarize positive- and negative-control separation when its assumptions are appropriate. A value of 0.5 or greater is a useful conventional target, not a substitute for run-specific signal-window, drift, edge-effect, reference-ligand, and curve-quality review [28].

21.3 Low-efficacy reporting

Percentage coefficient of variation becomes unstable when Emax approaches zero. Low-efficacy responses should be reported using absolute uncertainty, confidence intervals, assay noise, lower limits of response quantification, and replicate-level curves. An Emax estimate should not be ranked when the confidence interval spans inactivity and the effect is not reproduced orthogonally.

21.4 Donor and batch models

For human olfactory studies, donor is the primary biological replicate. Mixed-effects or hierarchical models should account for donor, batch, plate, and repeated measures. Sex, age, phenotype, duration, inflammatory status, collection site, handling time, and culture quality should be prespecified covariates or stratification variables when sample size permits.

21.5 Reproducibility before advancement

Advancement-critical findings require an internal confirmation and, before candidate nomination, replication by an independent laboratory or a clearly independent platform. Protocol deviations, negative data, near-threshold results, and failed transfers must remain in the controlled record.

22. Vendor-Ready CRO Framework

22.1 Vendor qualification rules

• Paid reference-ligand feasibility run before full specialized signaling work.

• Written instrument, construct, cell-line, passage, receptor-expression, and analysis details.

• Raw data, plate maps, chromatograms, images, scripts, metadata, exclusions, deviations, and quality-control results.

• No unauthorized compound substitution, structure re-registration, assay substitution, data transformation, or pooling.

• Predefined acceptance criteria, milestone-based releases, audit rights, reserve samples, and transfer obligations.

• Failure classification as assay, compound, model, supply, or vendor-execution failure before portfolio action.

22.2 Bioanalytical recovery requirement

The CRO package must quantify parent exposure at time zero and endpoint in every decision-driving matrix. For cellular systems, the method should recover and report supernatant, cell-associated, plate-rinse or adsorbed, precipitated, and major metabolite fractions when technically feasible. Preferred total recovery is 80% to 120%; recovery below 70% restricts quantitative interpretation, and below 60% invalidates the run unless a predefined alternative criterion is scientifically justified.

23. Regulatory and Intellectual-Property Position

23.1 Structural similarity does not decide FDA approval

A molecule is not barred from FDA approval merely because it resembles psilocin, and a structurally distinct molecule is not favored merely because it is different. Approval depends on the total evidence for identity, manufacturing quality, pharmacology, safety, efficacy, benefit-risk, labeling, and control of foreseeable risks. Structural distance may affect novelty, scheduling, metabolism, pharmacology, and development strategy, but it does not replace the evidence required for approval.

23.2 New molecular entity and scheduling are separate analyses

The NR-100 platform may offer a distinct chemical identity and potentially stronger patent positioning than a direct psilocin analog. Whether it qualifies as a new molecular entity, falls within an existing patent claim, or triggers controlled-substance treatment requires formal legal and regulatory review. Psychoactive pharmacology and abuse potential may influence scheduling even when the structure is not a classical psychedelic.

23.3 Patent workflow

1. Freeze authoritative structures and all inventorship records before broad disclosure.

2. Perform exact, substructure, similarity, scaffold, Markush, salt, stereochemistry, formulation, method-of-use, and target searches across patents and non-patent literature.

3. Separate patentability from freedom-to-operate analysis.

4. Document the conception and experimental support for each claimed embodiment.

5. Use a registered patent professional with pharmaceutical and new-chemical-entity experience before filing or public novelty claims.

23.4 Regulatory sequence

The near-term program is preclinical discovery. A future regulatory path would require pre-IND engagement, controlled CMC development, GLP toxicology as appropriate, safety pharmacology, genotoxicity, reproductive and chronic plans matched to indication and dosing, abuse-potential assessment when indicated, an IND, phased clinical trials, and a marketing application supported by substantial evidence. The exact package depends on the compound, route, duration, indication, and regulatory feedback.

24. Supportable Claims and Communications

24.1 Publication rule

Public materials must distinguish calculated properties, modeled hypotheses, preclinical observations, and clinical evidence. Docking must never be presented as receptor activity. Animal head-twitch results must never be presented as proof of human non-hallucinogenic action. Self-published proposals must be labeled as hypotheses and must not imply peer review or experimental validation.

25. Limitations

• NR-101 through NR-106 have no experimental chemical or biological validation.

• The written SMILES records and calculated formulas require independent structure verification.

• The naphthalene scaffold may be inactive, antagonistic, broadly promiscuous, poorly soluble, or toxic.

• Morpholine may reduce target affinity, create insufficient intracellular exposure, or still permit lysosomal trapping and phospholipidosis.

• Calculated pKa, logD, trapping, and metabolic predictions may not match measured material behavior.

• Ideal ionization and lysosomal-partition equations omit transport, binding, membrane potential, organelle volume, metabolism, and nonequilibrium behavior.

• Candidate property values without controlled structure hashes, model versions, input assumptions, uncertainty, and dates are not authoritative records.

• The free naphthol may undergo rapid glucuronidation or sulfation, but the rate and consequence for each NR-100 molecule are unknown.

• Low 5-HT2A efficacy and reduced Gi activity may not reliably predict human subjective effects.

• 5-HT1A and 5-HT1B contributions are species-, circuit-, sex-, stress-, and endpoint-dependent.

• Human olfactory organoids and epithelial cultures are developing models with incomplete clinical predictivity.

• An apparent olfactory phenotype may not translate into restoration of human smell perception.

• Intranasal delivery can create local exposure but also local toxicity, clearance, swallowing, and variable deposition.

• Patent novelty, freedom to operate, regulatory classification, manufacturability, and commercial viability remain unknown.

26. Milestones and Decision Charter

26.1 Immediate next action

26.2 Program kill rule

The NR-100 naphthalene program shall be paused when qualified assays show no reproducible target-relevant activity across the initial cohort after inadequate recovery, precipitation, adsorption, aggregation, cytotoxicity, identity error, and platform failure have been excluded. Broad backup chemistry is prohibited as a default rescue. A restart requires a new evidence-based pharmacophore rationale and formal portfolio approval.

27. Conclusion

NeuroRasool Version 4.3 retains the NR-100 hydrazine-free discovery platform and NR-101 as the provisional first test article. NR-101 is a substituted naphthol containing an isopropyl steric probe and a three-carbon tether terminating at the morpholine nitrogen. NR-102 through NR-106 provide matched tests of basicity, polarity, steric shape, electronic substitution, and linker geometry. No structural change or modeled property in this update is presented as proof of safety.

The redesigned molecule is not presented as a finished non-hallucinogenic drug. Its purpose is to test whether the NeuroRasool chemical concept has legitimate pharmacology and whether that pharmacology produces direct olfactory benefit without unacceptable safety or psychoactive liabilities. Low 5-HT2A efficacy and reduced Gi recruitment are evidence-based counter-screen hypotheses, not validated predictors of human subjective effects.

The decisive outcome of v4.3 may be advancement, redesign, reclassification, or termination. The first practical step is a paired structure-control and assay-readiness package governed by a shared dependency register and joint release rule. Qualified material must then earn advancement through measured pKa/logD, phenolic conjugation, hepatocyte clearance, metabolite, trapping, recovery, pharmacology, olfactory-function, and safety evidence.

Appendix A. Candidate Records

NR-101: Primary balance hypothesis

NR-102: Polar strong-base comparator

NR-103: Acyclic strong-base comparator

NR-104: Aromatic steric comparator

NR-105: Lower-lipophilicity electronic comparator

NR-106: Linker-geometry comparator

Appendix B. Assay Acceptance Matrix

Appendix C. RFP Deliverables

• Signed scope, methods, acceptance criteria, schedule, responsibilities, and change-control procedure.

• Controlled compound manifest with structure files, identifiers, hashes, batch, purity, storage, and chain of custody.

• Assay qualification report with constructs, cell provenance, controls, reagent lots, raw data, analysis model, and deviations.

• Exposure report with calibration, QC, chromatograms, recovery, stability, mass balance, exclusions, and corrective actions.

• Per-compound report containing full curves, replicate-level estimates, confidence intervals, cytotoxicity, and raw plate data.

• Human-model report containing tissue governance, donor metadata, identity, culture qualification, exclusion rules, raw images, functional data, and hierarchical analyses.

• Integrated decision report mapping every result to the version-controlled gate criteria.

• Return or controlled disposition of reserve material, biospecimens, data, and proprietary records at project completion.

Appendix D. Equation Reference and Model Boundaries

The following equations are embedded as editable Word mathematics and collected here for technical review. Every reported result must retain the input values and units, biological matrix or compartment, source-data identifier, calculation method or software and version, analyst, date, assumptions, uncertainty or sensitivity analysis, and linkage to the applicable controlled record. Model outputs organize decisions; they do not establish safety, efficacy, target engagement, or human exposure.

D.1 Neutral Fraction for an Equilibrium Monoprotic Base

Use: estimates the uncharged fraction at a specified pH. Boundary: assumes one relevant basic ionization, equilibrium conditions, and compatible pKa and pH conventions. It does not measure permeability, membrane binding, transport, solubility, or intracellular free concentration.

D.2 Idealized Lysosomal pH-Partition Ratio

Use: compares idealized total concentration between lysosomal and cytosolic pH conditions for a freely diffusible monoprotic base. Boundary: excludes membrane potential, organelle volume, binding, active transport, metabolism, precipitation, saturation, and kinetic limitations. Report pHlys, pHcyt, pKa, temperature, ionic conditions, and sensitivity ranges.

D.3 Cellular-System Mass Balance

Use: reconciles dosed material across operationally defined fractions. Boundary: every term must use compatible units and sampling times. Cell-associated material is not automatically target-accessible, and metabolites must be reported individually or converted using a declared parent-equivalent basis.

D.4 Functional Rescue Index

Use: expresses treatment-associated recovery relative to the prespecified healthy-control and injured-control contrast. Boundary: the denominator must be nonzero, stable, and biologically meaningful. Report the raw treated, injured, and healthy-control effects, donor-level uncertainty, assay dynamic range, and any values outside 0% to 100%. Do not truncate overshoot or worsening without a prespecified rule.

D.5 Operational 5-HT2B Margin

Use: supports exposure-based interpretation of a defined 5-HT2B functional effect. Boundary: numerator and denominator must use compatible concentration units and clearly defined effect and exposure metrics. Incorporate unbound exposure, duration, accumulation, uncertainty, and active metabolites. The ratio is a prioritization tool and never proof of cardiovascular safety.

D.6 Four-Parameter Logistic Curve Model

Use: estimates Bottom, Top, log EC50, and HillSlope when the concentration-response data support both asymptotes. X is the logarithm of concentration on the same base and unit convention used for log EC50. Boundary: report constraints, weighting, convergence, residuals, confidence intervals, excluded observations, and justification for any alternative model.

D.7 Z-Prime Assay-Quality Metric

Use: summarizes separation and variability of positive and negative controls for an applicable screening run. Boundary: control populations must be stable and representative, with drift and edge effects examined separately. A Z-prime value of at least 0.5 is a conventional target, not a universal acceptance rule or a substitute for full run review [28].

D.8 Minimum Provenance Record

References

1. Vargas MV, Dunlap LE, Dong C, et al. Psychedelics promote neuroplasticity through the activation of intracellular 5-HT2A receptors. Science. 2023;379:700-706. doi:10.1126/science.adf0435.
2. Wallach J, Cao AB, Calkins MM, et al. Identification of 5-HT2A receptor signaling pathways associated with psychedelic potential. Nature Communications. 2023;14:8221. doi:10.1038/s41467-023-44016-1.
3. Xu Z, Wang H, Yu J, et al. Psychedelics elicit their effects by 5-HT2A receptor-mediated Gi signalling. Nature. 2026;651:829-837. doi:10.1038/s41586-025-10061-7.
4. Ippolito A, Vasudevan S, Hurley S, et al. Evidence that 5-HT2A receptor signalling efficacy and not biased agonism differentiates serotonergic psychedelic from non-psychedelic drugs. British Journal of Pharmacology. 2026;183:4058-4071. Epub 2025. doi:10.1111/bph.70109.
5. Agrawal R, Gillie D, Mungenast A, et al. Zalsupindole is a nondissociative, nonhallucinogenic neuroplastogen with therapeutic effects comparable to ketamine and psychedelics. ACS Chemical Neuroscience. 2025;16. doi:10.1021/acschemneuro.5c00667.
6. Cameron LP, Tombari RJ, Lu J, et al. A non-hallucinogenic psychedelic analogue with therapeutic potential. Nature. 2021;589:474-479. doi:10.1038/s41586-020-3008-z.
7. Puigseslloses P, et al. Neuropharmacology of halogenated DMT analogs: psychoplastogenic and antidepressant properties of 5-Br-DMT, a psychedelic derivative with low hallucinogenic potential. Molecular Psychiatry. Published October 21, 2025. Article s41380-025-03308-2.
8. Fleury S, Nautiyal KM. The serotonin 1B receptor is required for some of the behavioral effects of psilocybin in mice. Molecular Psychiatry. Published December 20, 2025; volume 31:2685-2697. Article s41380-025-03387-1.
9. Sekssaoui M, et al. Antidepressant-like effects of psychedelics in a chronic despair mouse model: is the 5-HT2A receptor the unique player? Neuropsychopharmacology. 2024.
10. Finlay JB, Brann DH, Abi Hachem R, et al. Persistent post-COVID-19 smell loss is associated with immune cell infiltration and altered gene expression in olfactory epithelium. Science Translational Medicine. 2022;14:eadd0484. doi:10.1126/scitranslmed.add0484.
11. Kikuta S, Nagayama S, Hasegawa-Ishii S. Structures and functions of the normal and injured human olfactory epithelium. Frontiers in Neural Circuits. 2024;18:1406218. doi:10.3389/fncir.2024.1406218.
12. Kim S, Finlay JB, Ko T, Goldstein BJ. Long-term olfactory loss post-COVID-19: pathobiology and potential therapeutic strategies. World Journal of Otorhinolaryngology Head and Neck Surgery. 2024;10:148-155. doi:10.1002/wjo2.165.
13. Liu J, Zhang Y, Yu Y. Establishment of nasal and olfactory epithelium organoids for unveiling mechanisms of tissue regeneration and pathogenesis of nasal diseases. Cellular and Molecular Life Sciences. 2025;82:33. doi:10.1007/s00018-024-05557-w.
14. Gameiro JG, Virtanen J, Lee L, Schwob JE, Fornazieri MA, Lin B. Protocol for culturing olfactory epithelium organoids supporting neuronal differentiation. STAR Protocols. 2025;6:104211. doi:10.1016/j.xpro.2025.104211.
15. Liu J, Qi J, Jiang N, et al. Olfactory epithelium organoid models identify Ddit3 as a potential therapeutic target against inflammation-related olfactory sensory neuronal loss and functional deficit. International Journal of Biological Sciences. 2026;22:4025-4042. doi:10.7150/ijbs.129192.
16. Dumotier BM, Urban L. Preclinical mitigation of 5-HT2B agonism-related cardiac valvulopathy revisited. Journal of Pharmacological and Toxicological Methods. 2024;128:107542.
17. Kalgutkar AS. Designing around structural alerts in drug discovery. Journal of Medicinal Chemistry. 2020;63:6276-6302. doi:10.1021/acs.jmedchem.9b00917.
18. Breiden B, Sandhoff K. Emerging mechanisms of drug-induced phospholipidosis. Biological Chemistry. 2020;401:31-46. doi:10.1515/hsz-2019-0270.
19. Waring MJ, Johnstone C. A quantitative assessment of hERG liability as a function of lipophilicity. Bioorganic and Medicinal Chemistry Letters. 2007;17:1759-1764.
20. Pottie E, Stove CP. In vitro assays for the functional characterization of psychedelic substances at the serotonin 5-HT2A receptor. Journal of Neurochemistry. 2022;162:39-59. doi:10.1111/jnc.15570.
21. International Council for Harmonisation. ICH M7(R2): Assessment and Control of DNA Reactive Mutagenic Impurities in Pharmaceuticals to Limit Potential Carcinogenic Risk. 2023.
22. International Council for Harmonisation. ICH S2(R1): Guidance on Genotoxicity Testing and Data Interpretation for Pharmaceuticals Intended for Human Use. 2011.
23. International Council for Harmonisation. ICH M10: Bioanalytical Method Validation and Study Sample Analysis. 2022.
24. U.S. Food and Drug Administration. General Considerations for the Use of New Approach Methodologies in Drug Development. Draft Guidance for Industry. March 2026.
25. National Cancer Institute. NCI Best Practices for Biospecimen Resources. Fourth Edition. 2026.
26. Moore HM, Kelly AB, Jewell SD, et al. Biospecimen reporting for improved study quality. Cancer Cytopathology. 2011;119:92-101. doi:10.1002/cncy.20147.
27. International Council for Harmonisation. E14/S7B Questions and Answers: Clinical and Nonclinical Evaluation of QT/QTc Interval Prolongation and Proarrhythmic Potential. 2022.
28. Zhang JH, Chung TDY, Oldenburg KR. A simple statistical parameter for use in evaluation and validation of high throughput screening assays. Journal of Biomolecular Screening. 1999;4:67-73.
29. Schmitt MV, Lienau P, Fricker G, Reichel A. Quantitation of lysosomal trapping of basic lipophilic compounds using in vitro assays and in silico predictions based on the determination of the full pH profile of the endo-/lysosomal system in rat hepatocytes. Drug Metabolism and Disposition. 2019;47:49-57. doi:10.1124/dmd.118.084541.
30. Pacifici GM, Franchi M, Giuliani L, Rane A. Development of the glucuronyltransferase and sulphotransferase towards 2-naphthol in human fetus. Developmental Pharmacology and Therapeutics. 1989;14:108-114. PMID:2517106.
31. Madden MB, et al. Serotonin and psilocybin activate 5-HT1B receptors to suppress cortical signaling through the claustrum. Nature Communications. 2025;16:7733. doi:10.1038/s41467-025-62980-8.

Document Control