## One-Page Proposal — Emergent Ventures Africa **Project:** Conjunctive Consolidation Threshold (CCT): Fixing Addiction Pharmacology at Its Causal Root **Applicant:** Eniola Ayodele Olutogun | OIQB, Lagos | zyco.org | ORCID: 0009-0001-9272-6735 --- Addiction pharmacology has failed for fifty years — not because the problem is intractable, but because the field has been intervening at the wrong point in the causal chain. We treat withdrawal, cravings, and relapse. We leave untouched the mechanism that makes addiction self-perpetuating: the pathological encoding of reward memories. Every high creates a consolidated memory trace that survives detoxification. Relapse rates above 60% are not a willpower failure. They are a scientific failure to intervene before the memory is written. The Conjunctive Consolidation Threshold (CCT) model is my attempt to correct this from first principles. It is a tripartite pharmacological framework — co-targeting NMDA receptors, dopamine D1/D5 receptors, and AMPA receptor trafficking — designed to prevent reward-memory consolidation during the critical post-exposure window. The framework predicts that simultaneous modulation of all three systems reduces encoding probability from 0.855 to 0.122 (an 85.8% reduction), with super-additive interaction effects of +12.8 percentage points. All five pre-registered hypotheses confirmed across ODE/RK45 simulation and Bayesian MCMC validation. I have done this work alone. The Olutogun Institute of Quantitative Biosciences (OIQB) is a one-person computational lab operating out of Lagos, Nigeria. Three sole-authored preprints, a Bayesian clinical trial architecture, and a review article currently under review at *Neuroscience & Biobehavioral Reviews*. Kent Berridge at Michigan, Samuel Gershman at Harvard (arXiv endorsement), Nathaniel Daw at Princeton, and Marcelo Mattar at NYU found the work credible enough to engage with substantively. The Africa dimension is structural, not rhetorical. Sub-Saharan Africa has no computational neuropharmacology infrastructure — no academic groups doing mechanistic CNS modelling, no institutional pipelines, no accessible HPC clusters for independent researchers. The tools I am building alongside CCT — IMPRINT for addiction-liability screening, TOPOLOGIX for topological drug-protein interaction analysis, GATE for BCI safety evaluation — are designed to seed that infrastructure from the ground up. Emergent Ventures funds what institutional review committees will not touch: genuinely novel bets by people operating outside sanctioned structures. CCT is not a proposal to run an incremental clinical study. It is a claim that the entire therapeutic strategy in addiction is aimed at the wrong biological target. I am requesting $50,000 to support twelve months of full-time computational development, experimental partnership brokering with wet-lab collaborators, and provisional patent prosecution for the CCT core architecture — converting a validated theoretical framework into a clinically stageable intervention. There is a Nigerian pharmacist building computational neuroscience from scratch, without a PhD, without a university, without a lab. That is the kind of bet Emergent Ventures exists to make. *(~370 words)* --- ## Substack Post — Evidence of Work **Title: The Memory Is the Disease — Why Addiction Treatment Has Been Hitting the Wrong Target** *Posted at zyco.org / Substack — [your handle]* --- Here is a fact that should bother you more than it does: the majority of people treated for opioid or stimulant addiction relapse within a year. Not most. The majority. And this has been true for decades, across every approved pharmacotherapy we have. The standard explanation is that addiction is a chronic brain disease with strong biological underpinnings, and so relapse is expected, like hypertension coming back when medication stops. This explanation is technically defensible and strategically disastrous, because it quietly absolves the science of having to explain why the biology keeps winning. I think there is a more uncomfortable explanation: we have been treating the wrong biological target. --- When someone uses a substance compulsively, two things happen simultaneously. The first is the pharmacological event everyone focuses on — dopamine surge, receptor activation, reward signalling. The second is something neuroscience has studied extensively but pharmacology has largely ignored: memory consolidation. The brain is not just experiencing pleasure. It is *learning* from it. It is encoding the context, the cues, the sequence of events, into long-term memory via a coordinated process involving NMDA receptors, dopamine D1/D5 receptors, and AMPA receptor trafficking. That memory does not disappear with detoxification. It is why a recovered heroin user can walk past a specific street corner and feel craving flood back twenty years later. The memory is structurally intact. We never touched it. Current pharmacotherapies — naltrexone, methadone, buprenorphine, varenicline — work on the reward signal itself. They reduce the high, reduce craving, reduce withdrawal symptoms. None of them were designed to intervene in the window during which the reward experience gets *written into long-term memory*. That window exists, it is measurable, and it has a mechanistic threshold. This is what I have been working on. --- The Conjunctive Consolidation Threshold (CCT) model proposes that reward-memory encoding requires conjunctive activation across three molecular systems — NMDA, D1/D5, and AMPA trafficking — simultaneously. Disrupt any one system partially and the memory is attenuated. Disrupt all three together, in the correct pharmacological relationship, and encoding probability collapses. The model predicts an 85.8% reduction in encoding probability under full tripartite modulation, with super-additive interaction effects that no single-target or dual-target approach can replicate. I should be transparent about what this currently is: a computational framework. Three preprints, ODE/RK45 differential equation modelling, Bayesian MCMC validation, a proposed clinical trial architecture. The experimental work in humans is downstream. But the mechanistic logic is not speculative — it is built from the existing literature on synaptic consolidation, fear memory extinction, and LTP/LTD dynamics, recombined to ask a question that the field has not asked: what is the minimum pharmacological condition necessary to prevent this specific encoding event? That question has a mathematical answer. The threshold is real. And if the threshold is real, then there exists a treatment window — shortly after exposure — during which a correctly timed intervention could prevent the addiction memory from forming in the first place. --- I am based in Lagos. I work independently, without institutional affiliation, under the aegis of OIQB (the Olutogun Institute of Quantitative Biosciences), which is a name I gave to my research practice because the work needed a home. I have a pharmacy degree from the University of Ibadan and I am a licensed pharmacist. I have no PhD, no lab, no university salary. What I have is a computational pipeline, three preprints, colleagues at Michigan and Harvard and Princeton who took the work seriously, and a deep conviction that the right intervention in addiction is upstream of where medicine is currently looking. Sub-Saharan Africa has no computational neuropharmacology infrastructure. No groups doing mechanistic CNS modelling at the level I am describing. No pipelines connecting African pharmacological data to global research ecosystems. Part of what I am trying to do — in parallel with CCT — is show that this infrastructure can exist here, built from first principles, by people who have reasons to care about getting it right. The memory is the disease. Until pharmacology treats it as such, we will keep achieving partial victories against a target that keeps regenerating. *(~680 words — edit to platform length preference; aim for 500–700 for Substack)* --- ## 1-Page CV --- # ENIOLA AYODELE OLUTOGUN Lagos, Nigeria | work@[email] | zyco.org | github.com/AmunRaPtah ORCID: 0009-0001-9272-6735 | PCN Licence: [number] --- ### EDUCATION **B.Pharm (Honours)**, University of Ibadan — 2014–2021 CGPA 5.1/7.0 · 2:1 Upper Division · German equivalent: 1.9 *MSc application submitted: Medical University of Graz, Austria (Oct 2026 intake)* --- ### INDEPENDENT RESEARCH — OIQB, Lagos (2025–present) **Conjunctive Consolidation Threshold (CCT) Model** *(sole author)* Tripartite pharmacological framework for reward-memory encoding prevention in addiction. ODE/RK45 + Bayesian MCMC validation; encoding probability reduced 0.855→0.122 (85.8%); super-additivity +12.8 pp; H1–H5 confirmed. - Preprint 1: Foundational CCT paper — OSF 10.17605/OSF.IO/KG7B5 - Preprint 2: Formal mathematical specification — OSF 10.17605/OSF.IO/EMY4U - Preprint 3: Bayesian population dynamics + clinical trial architecture — Zenodo 10.5281/zenodo.20492472 - Review article under review, *Neuroscience & Biobehavioral Reviews* - Co-authored paper under review, *Alcohol* (Elsevier) - Provisional patent on CCT core architecture (Q3 2026) **Computational Platforms Built:** - **IMPRINT** — addiction-liability screening platform - **TOPOLOGIX** — TDA for drug-protein interaction (persistent homology, bipartite simplicial complexes, hERG cardiotoxicity MVP) - **GATE** — BCI neural-stimulation safety evaluation (Apache 2.0) --- ### COLLABORATORS / EXTERNAL ENGAGEMENT Kent Berridge (Univ. Michigan) · Samuel Gershman (Harvard; arXiv endorsement) · Nathaniel Daw (Princeton) · Marcelo Mattar (NYU) --- ### EMPLOYMENT **National Product Manager** — Synthcare, Lagos (Mar 2026–present) **Clinical Pharmacist** — Ramset Pharmacy (Jan–Mar 2026) **Research Assistant** — CDDDP, Lagos (NMDA/insulin receptor docking) **Bioinformatics Researcher** — GHRU-GSAR (AMR genomics; surveillance pipeline) --- ### TECHNICAL SKILLS **Computational:** Python (scipy, numpy, PyMC/MCMC, pandas), R, ODE/RK45, TDA (Ripser, Gudhi), NEURON/Brian2, AlphaFold, RDKit, ADMET/QSAR, GROMACS, AutoDock **Infrastructure:** Nextflow, SLURM/HPC, Supabase/Postgres, JavaScript/Node.js **Domain:** Computational neuropharmacology, molecular dynamics, population PK/PD, Bayesian inference --- ### SELECTED TRAINING / RECOGNITION *(Add any relevant workshops, hackathon placements, scholarships, or awards here)* --- ## Checklist - ☐ Complete online Emergent Ventures application form at mercatus.org/emergent-ventures - ☐ Paste or upload the one-page proposal (~370 words) into the application text field - ☐ Confirm all three OSF/Zenodo preprint DOIs are publicly accessible and resolving - ☐ Publish (or schedule) the Substack post so it is live and linkable at submission time - ☐ Include Substack URL as evidence of intellectual work in the application - ☐ Add CV as attachment or linked PDF (confirm whether EV accepts attachments — if not, paste inline) - ☐ Fill in PCN licence number on CV - ☐ Fill in contact email on CV - ☐ Add provisional patent application number to CV/proposal once filed (Q3 2026 target — note if pending) - ☐ Confirm ORCID profile is public and lists all three preprints - ☐ Add GitHub repo links for GATE (Apache 2.0 already), IMPRINT, and TOPOLOGIX if repositories are public - ☐ Verify collaborator names and affiliations are current (Daw → Princeton, Mattar → NYU, Gershman → Harvard, Berridge → Michigan) - ☐ Confirm arXiv endorsement from Gershman is documented / linkable - ☐ Review EV application page for any updated field requirements before submitting --- ## Editor Notes - **Patent filing status needs precision.** The proposal says "provisional patent Q3 2026" but as of submission this may be unfiled. Either remove the reference or replace with "patent application in preparation" — overstating IP status is a credibility risk if Tyler asks directly. - **The Substack must exist before submission.** EV looks for evidence of intellectual presence; linking to a post that does not yet exist (or a private draft) undermines the application. If there is no active Substack, publish at least one substantive post first and link it. Alternatively, link to the OSF preprints as the primary evidence and omit the Substack reference if the channel does not yet have public content. - **The proposal needs Eniola's authentic register.** The draft is written to sound credible and ambitious — but Tyler Cowen rewards intellectual personality, not polish. Read the proposal aloud: if it sounds like a grant committee wrote it, add a sentence or two in genuine first-person voice. The paragraph beginning "I have done this work alone" is the strongest; consider leading with that energy earlier. - **Budget justification is absent.** The proposal requests $50,000 but gives no breakdown (compute costs, travel for experimental partnerships, patent prosecution fees, runway). EV does not always require this, but a single sentence — "split across HPC compute access, wet-lab collaboration travel, and patent costs" — anticipates an obvious follow-up question from a careful reader.