Draft — Pharmacology 2026

MOTIVATION LETTER

The Conjunctive Consolidation Threshold model, or CCT, is a tripartite pharmacological framework for preventing reward-memory encoding in addiction. It couples three axes, dopaminergic reward prediction error, NMDAR-dependent long-term potentiation, and affective contrast, into a single system of ordinary differential equations solved with RK45. The model was calibrated using Bayesian MCMC with PyMC's DEMetropolisZ sampler across 14 free parameters, with priors elicited from a systematic screen of 1,847 records from the literature. All five pre-registered hypotheses, H1 through H5, were confirmed, and the posterior shows super-additivity of 13 to 22 percentage points across model versions. This work is currently under review at three peer-reviewed journals: International Addiction Review, Progress in Neuro-Psychopharmacology and Biological Psychiatry, and Neuroscience and Biobehavioral Reviews.

I am applying to Pharmacology 2026 because the British Pharmacological Society explicitly supports the next generation of researchers, and my work offers a quantitative, systems-level perspective that complements the experimental presentations typical of this meeting. The CCT model makes testable predictions about how combined pharmacological interventions, targeting dopamine D2 receptors, NMDARs, and affective valence circuits, could block the consolidation of drug-associated memories. These predictions are specific enough to guide dose-finding and combination-strategy studies in animal models. For example, the model predicts that simultaneous sub-threshold modulation across all three axes produces a supra-linear effect that no single-axis intervention achieves. That is a directly falsifiable claim a preclinical lab could test within a year.

My training as a pharmacist at the University of Ibadan, graduating with a 2:1 Upper Division and a German equivalent of 1.9, grounds this computational work in clinical pharmacology. I have since built a receptor-to-behavior simulation engine called neurocascade, which couples pharmacokinetics to receptor binding to Wilson-Cowan circuit dynamics. That engine has 62 passing tests across three literature-calibrated receptor systems: mu-opioid, D2 dopamine, and GABA-A. The CCT model is the disease-focused application of the same methodological stack.

What distinguishes this submission is the pre-registration and the Bayesian calibration. The hypotheses were registered before any fitting was performed. The priors are documented and traceable to a specific literature screen. The posterior distributions are reported, not just point estimates. This is the standard of rigor that pharmacological research should adopt as computational models become more common in drug discovery, and I want to present this standard to the BPS community at Pharmacology 2026.

The conference is also an opportunity to establish collaborations. I have received endorsements from Kent Berridge at Michigan, Samuel Gershman at Harvard, Nathaniel Daw at Princeton, and Marcelo Mattar at NYU. I am seeking experimental partners who can test the CCT model's predictions in vivo. Pharmacology 2026 is the right venue to find them.

RESEARCH STATEMENT

The CCT model addresses a core question in addiction pharmacology: can we prevent the formation of drug-associated memories before they become compulsive? Current pharmacotherapies for addiction, such as naltrexone and acamprosate, act on single targets and show modest effect sizes. The CCT model argues that reward-memory consolidation is a multi-axis process, and that effective prevention requires simultaneous modulation of dopaminergic reward prediction error, NMDAR-dependent plasticity, and affective contrast. The model formalizes this as a system of coupled ODEs, where each axis has its own dynamics and the axes interact multiplicatively at a consolidation threshold.

The model was built in three stages. First, I conducted a systematic literature screen of 1,847 records to elicit priors for all 14 free parameters. These priors are derived from published dose-response curves, receptor binding affinities, and electrophysiological data. Second, I implemented the ODE system in Python using scipy's RK45 solver and calibrated it against behavioral data from conditioned place preference and self-administration studies using PyMC's DEMetropolisZ sampler. Third, I pre-registered five hypotheses, H1 through H5, covering the model's predictions about single-axis versus multi-axis interventions, dose-response relationships, and temporal windows of consolidation. All five hypotheses were confirmed in the posterior analysis.

The key result is super-additivity. When all three axes are modulated at sub-threshold levels, the model predicts a 13 to 22 percentage point increase in consolidation prevention compared to the sum of individual effects. This is a multiplicative interaction at the threshold, not an additive effect. The clinical implication is that low-dose combination therapy, with reduced side-effect burden, could outperform high-dose monotherapy. The model also predicts a critical time window, approximately 6 to 12 hours post-exposure, during which consolidation can be disrupted. After that window, the memory trace is stable and resistant to intervention.

The CCT model is currently under review at three journals: International Addiction Review, Progress in Neuro-Psychopharmacology and Biological Psychiatry, and Neuroscience and Biobehavioral Reviews. A co-authored paper on related alcohol pharmacology is under review at Alcohol, published by Elsevier. The model has not yet been tested in vivo, and I am explicit about that limitation. What the model offers is a rigorous, falsifiable framework that can prioritize which combinations to test and at what doses.

My broader methodological work supports this research. I have developed neurocascade, a receptor-to-behavior simulation engine that couples pharmacokinetics to receptor binding to Wilson-Cowan circuit dynamics. The engine is Bayesian-calibrated and has 62 passing tests across mu-opioid, D2 dopamine, and GABA-A systems. I have also worked on protein-language-model approaches to drug resistance prediction, achieving an AUROC of 0.804 on the Platinum benchmark, and on topological data analysis for cardiotoxicity prediction, where I published a pre-registered replication showing that topological features do not beat a plain descriptor baseline. That negative result was reported directly, not reframed.

For Pharmacology 2026, I propose to present the CCT model's architecture, its Bayesian calibration, and its testable predictions. The presentation would include the posterior distributions, the super-additivity analysis, and a concrete experimental protocol for testing the model's predictions in a preclinical setting. I am also prepared to discuss the model's limitations, including the illustrative nature of the circuit-layer parameters in neurocascade and the need for real behavioral data to refine those parameters.

The BPS mission to support early-career researchers aligns with my current stage. I am enrolled in the M.Sc. Digital Health program at the Hasso Plattner Institute and the University of Potsdam, starting in the Winter Semester of 2026/27. I am an independent researcher with a track record of pre-registered, reproducible computational work. Pharmacology 2026 is an opportunity to present that work to a pharmacological audience, receive feedback from experimentalists, and establish the collaborations needed to move the CCT model from simulation to the bench.

EDITOR NOTES

- Framing choice: This submission leads with the CCT model, not TOPOLOGIX or the cardiotoxicity topology study. Rationale: CCT is the only research line that directly addresses a pharmacological question, addiction pharmacotherapy, and it is the line with the most rigorous validation (pre-registered hypotheses, Bayesian calibration, three journals under review). The cardiotoxicity study is a negative result and is mentioned only as evidence of methodological rigor, not as a current research line. TOPOLOGIX is omitted because it is a protein-ML project with no direct pharmacology angle for this conference.

- Eligibility risk: The programme page lists no eligibility conditions, but the applicant is a Nigerian national enrolled in a German M.Sc. program starting Winter 2026/27. If the conference requires affiliation with a UK institution or a BPS membership, this could be a barrier. Verify BPS membership requirements and whether independent researchers can present without institutional sponsorship.

- Verification needed: The three journals under review (IART, PNPBP, NBR) are named in the profile but the applicant should confirm the current review status before submitting. The co-authored paper at Alcohol should also be confirmed as under review. Do not state "under review" if the status has changed to accepted or rejected.

- Gap to fill: The applicant must insert personal details about why Pharmacology 2026 specifically, such as any prior attendance, specific sessions or speakers they want to engage with, and whether they plan to present a poster or oral talk. The programme page does not specify submission formats, so the applicant should check the website for abstract submission guidelines and deadlines.

- Honesty constraint: The neurocascade circuit-layer parameters are explicitly labeled illustrative pending real behavioral data. Do not claim these parameters are validated. The CCT model has not been tested in vivo; do not imply it has. The ergofluids project failed its first real-data gate and is not mentioned in this submission; do not include it.

MOTIVATION LETTER

The Conjunctive Consolidation Threshold model, or CCT, is a tripartite pharmacological framework for preventing reward-memory encoding in addiction. It couples three axes, dopaminergic reward prediction error, NMDAR-dependent long-term potentiation, and affective contrast, into a single system of ordinary differential equations solved with RK45. The model was calibrated using Bayesian MCMC with PyMC's DEMetropolisZ sampler across 14 free parameters, with priors elicited from a systematic screen of 1,847 records from the literature. All five pre-registered hypotheses, H1 through H5, were confirmed, and the posterior shows super-additivity of 13 to 22 percentage points across model versions. This work is currently under review at three peer-reviewed journals: International Addiction Review, Progress in Neuro-Psychopharmacology and Biological Psychiatry, and Neuroscience and Biobehavioral Reviews.

I am applying to Pharmacology 2026 because the British Pharmacological Society explicitly supports the next generation of researchers, and my work offers a quantitative, systems-level perspective that complements the experimental presentations typical of this meeting. The CCT model makes testable predictions about how combined pharmacological interventions, targeting dopamine D2 receptors, NMDARs, and affective valence circuits, could block the consolidation of drug-associated memories. These predictions are specific enough to guide dose-finding and combination-strategy studies in animal models. For example, the model predicts that simultaneous sub-threshold modulation across all three axes produces a supra-linear effect that no single-axis intervention achieves. That is a directly falsifiable claim a preclinical lab could test within a year.

My training as a pharmacist at the University of Ibadan, graduating with a 2:1 Upper Division and a German equivalent of 1.9, grounds this computational work in clinical pharmacology. I have since built a receptor-to-behavior simulation engine called neurocascade, which couples pharmacokinetics to receptor binding to Wilson-Cowan circuit dynamics. That engine has 62 passing tests across three literature-calibrated receptor systems: mu-opioid, D2 dopamine, and GABA-A. The CCT model is the disease-focused application of the same methodological stack.

What distinguishes this submission is the pre-registration and the Bayesian calibration. The hypotheses were registered before any fitting was performed. The priors are documented and traceable to a specific literature screen. The posterior distributions are reported, not just point estimates. This is the standard of rigor that pharmacological research should adopt as computational models become more common in drug discovery, and I want to present this standard to the BPS community at Pharmacology 2026.

The conference is also an opportunity to establish collaborations. I have received endorsements from Kent Berridge at Michigan, Samuel Gershman at Harvard, Nathaniel Daw at Princeton, and Marcelo Mattar at NYU. I am seeking experimental partners who can test the CCT model's predictions in vivo. Pharmacology 2026 is the right venue to find them.

RESEARCH STATEMENT

The CCT model addresses a core question in addiction pharmacology: can we prevent the formation of drug-associated memories before they become compulsive? Current pharmacotherapies for addiction, such as naltrexone and acamprosate, act on single targets and show modest effect sizes. The CCT model argues that reward-memory consolidation is a multi-axis process, and that effective prevention requires simultaneous modulation of dopaminergic reward prediction error, NMDAR-dependent plasticity, and affective contrast. The model formalizes this as a system of coupled ODEs, where each axis has its own dynamics and the axes interact multiplicatively at a consolidation threshold.

The model was built in three stages. First, I conducted a systematic literature screen of 1,847 records to elicit priors for all 14 free parameters. These priors are derived from published dose-response curves, receptor binding affinities, and electrophysiological data. Second, I implemented the ODE system in Python using scipy's RK45 solver and calibrated it against behavioral data from conditioned place preference and self-administration studies using PyMC's DEMetropolisZ sampler. Third, I pre-registered five hypotheses, H1 through H5, covering the model's predictions about single-axis versus multi-axis interventions, dose-response relationships, and temporal windows of consolidation. All five hypotheses were confirmed in the posterior analysis.

The key result is super-additivity. When all three axes are modulated at sub-threshold levels, the model predicts a 13 to 22 percentage point increase in consolidation prevention compared to the sum of individual effects. This is a multiplicative interaction at the threshold, not an additive effect. The clinical implication is that low-dose combination therapy, with reduced side-effect burden, could outperform high-dose monotherapy. The model also predicts a critical time window, approximately 6 to 12 hours post-exposure, during which consolidation can be disrupted. After that window, the memory trace is stable and resistant to intervention.

The CCT model is currently under review at three journals: International Addiction Review, Progress in Neuro-Psychopharmacology and Biological Psychiatry, and Neuroscience and Biobehavioral Reviews. A co-authored paper on related alcohol pharmacology is under review at Alcohol, published by Elsevier. The model has not yet been tested in vivo, and I am explicit about that limitation. What the model offers is a rigorous, falsifiable framework that can prioritize which combinations to test and at what doses.

My broader methodological work supports this research. I have developed neurocascade, a receptor-to-behavior simulation engine that couples pharmacokinetics to receptor binding to Wilson-Cowan circuit dynamics. The engine is Bayesian-calibrated and has 62 passing tests across mu-opioid, D2 dopamine, and GABA-A systems. I have also worked on protein-language-model approaches to drug resistance prediction, achieving an AUROC of 0.804 on the Platinum benchmark, and on topological data analysis for cardiotoxicity prediction, where I published a pre-registered replication showing that topological features do not beat a plain descriptor baseline. That negative result was reported directly, not reframed.

For Pharmacology 2026, I propose to present the CCT model's architecture, its Bayesian calibration, and its testable predictions. The presentation would include the posterior distributions, the super-additivity analysis, and a concrete experimental protocol for testing the model's predictions in a preclinical setting. I am also prepared to discuss the model's limitations, including the illustrative nature of the circuit-layer parameters in neurocascade and the need for real behavioral data to refine those parameters.

The BPS mission to support early-career researchers aligns with my current stage. I am enrolled in the M.Sc. Digital Health program at the Hasso Plattner Institute and the University of Potsdam, starting in the Winter Semester of 2026/27. I am an independent researcher with a track record of pre-registered, reproducible computational work. Pharmacology 2026 is an opportunity to present that work to a pharmacological audience, receive feedback from experimentalists, and establish the collaborations needed to move the CCT model from simulation to the bench.

EDITOR NOTES

- Framing choice: This submission leads with the CCT model, not TOPOLOGIX or the cardiotoxicity topology study. Rationale: CCT is the only research line that directly addresses a pharmacological question, addiction pharmacotherapy, and it is the line with the most rigorous validation (pre-registered hypotheses, Bayesian calibration, three journals under review). The cardiotoxicity study is a negative result and is mentioned only as evidence of methodological rigor, not as a current research line. TOPOLOGIX is omitted because it is a protein-ML project with no direct pharmacology angle for this conference.

- Eligibility risk: The programme page lists no eligibility conditions, but the applicant is a Nigerian national enrolled in a German M.Sc. program starting Winter 2026/27. If the conference requires affiliation with a UK institution or a BPS membership, this could be a barrier. Verify BPS membership requirements and whether independent researchers can present without institutional sponsorship.

- Verification needed: The three journals under review (IART, PNPBP, NBR) are named in the profile but the applicant should confirm the current review status before submitting. The co-authored paper at Alcohol should also be confirmed as under review. Do not state "under review" if the status has changed to accepted or rejected.

- Gap to fill: The applicant must insert personal details about why Pharmacology 2026 specifically, such as any prior attendance, specific sessions or speakers they want to engage with, and whether they plan to present a poster or oral talk. The programme page does not specify submission formats, so the applicant should check the website for abstract submission guidelines and deadlines.

- Honesty constraint: The neurocascade circuit-layer parameters are explicitly labeled illustrative pending real behavioral data. Do not claim these parameters are validated. The CCT model has not been tested in vivo; do not imply it has. The ergofluids project failed its first real-data gate and is not mentioned in this submission; do not include it.

AI Draft — Pharmacology 2026