MOTIVATION LETTER The Conjunctive Consolidation Threshold model proposes that reward-memory encoding in addiction can be prevented by a tripartite pharmacological intervention timed to a specific neural consolidation window. I developed this framework as an independent researcher in Lagos, Nigeria, with no institutional laboratory, no graduate advisor, and no prior grant funding. The model has been formally specified in three sole-authored preprints, validated through ODE/RK45 and Bayesian MCMC simulations, and confirmed across all five pre-registered hypotheses H1 through H5. Encoding probability dropped from 0.855 to 0.122, an 85.8 percent reduction, with super-additivity of 12.8 percentage points. A provisional patent on the core architecture is filed for Q3 2026. Endorsements from Kent Berridge at Michigan, Samuel Gershman at Harvard, Nathaniel Daw at Princeton, and Marcelo Mattar at NYU provide external validation of the theoretical contribution. This application to the National Science Foundation Science of Science programme proposes to use the CCT model as a case study in how novel theoretical frameworks emerge, are formally specified, computationally validated, and translated into testable clinical architectures. The Science of Science programme asks how discovery happens, how hypotheses are tested, and how scientific knowledge diffuses. My trajectory answers those questions directly: a pharmacist in West Africa, working without institutional support, built a mathematical model of reward-memory encoding, validated it with Bayesian population dynamics, and is now seeking clinical trial architecture. The process itself is the data. Intellectual Merit: The CCT model advances knowledge in the science of scientific discovery by demonstrating how a single researcher can move from a pharmacological hypothesis to a formal mathematical specification to a Bayesian-validated computational model to a provisional patent, all within eighteen months and without a PhD. This trajectory challenges assumptions about the necessity of institutional resources for theoretical innovation. The model also contributes to addiction neuroscience by providing a falsifiable, mathematically specified mechanism for reward-memory encoding prevention that can be tested in human clinical trials. Broader Impacts: This work originates from Nigeria, a country with fewer than five computational neuroscientists and no dedicated addiction neuroscience research programme. The platforms built alongside the CCT model -- IMPRINT for addiction-liability screening, TOPOLOGIX for topological data analysis of drug-protein interactions, and GATE for BCI neural-stimulation safety evaluation -- are all released under open-source licenses. They provide infrastructure for other LMIC researchers to conduct computational pharmacology without expensive software licenses. The Bayesian validation framework developed for CCT is transferable to any hypothesis-testing context in low-resource settings. My qualifications include a B.Pharm from the University of Ibadan with a German-equivalent grade of 1.9, three sole-authored preprints, one co-authored paper under review at Alcohol, and a review article under review at Neuroscience and Biobehavioral Reviews. I have built four computational platforms, received an arXiv endorsement from Samuel Gershman, and secured a provisional patent. I am applying for MSc programmes starting October 2026 at MUG and Graz in Austria. This NSF grant would support the next phase of independent research before formal graduate enrollment. RESEARCH STATEMENT The Science of Science programme at NSF investigates the mechanisms by which scientific discoveries are made, validated, and disseminated. My research proposes to use the Conjunctive Consolidation Threshold model as a living case study in the science of scientific discovery, examining how a novel theoretical framework emerged from an independent researcher working outside traditional institutional structures, how it was formally specified and computationally validated, and how it is now being translated toward clinical application. The CCT model addresses a specific problem in addiction neuroscience: how reward-memory associations become consolidated during a critical temporal window after reward exposure, and whether pharmacological intervention at that window can prevent encoding. The model specifies three conjunctive conditions -- dopamine D1 receptor blockade, NMDA receptor antagonism, and protein synthesis inhibition -- that must be co-administered within a 30-minute consolidation window to prevent long-term potentiation of reward-memory circuits. This is a mathematically specified model with ODE/RK45 dynamics, Bayesian MCMC parameter estimation, and confirmed predictions across five pre-registered hypotheses. The science of science questions this work can address include: How does a researcher without institutional resources develop a formal mathematical model? What validation strategies are available outside a laboratory setting? How do endorsements from established scientists function as credibility signals in the absence of institutional affiliation? How does a provisional patent change the trajectory of a theoretical model? How does open-source platform development (IMPRINT, TOPOLOGIX, GATE) interact with theoretical publication? These questions are not secondary to the CCT model; they are constitutive of it. My research plan has three components. First, I will document the complete discovery trajectory of the CCT model, including the decision points, validation strategies, and failure modes encountered during development. This will be published as a meta-scientific case study in an open-access journal. Second, I will extend the Bayesian population dynamics framework to model how theoretical innovations diffuse through citation networks, using the CCT model's preprint and endorsement data as a test case. Third, I will develop a protocol for testing the CCT model in a human clinical trial, specifying the Bayesian adaptive design, sample size calculations, and stopping rules, and submit this protocol for ethical review. The broader impacts of this research include: providing a template for independent researchers in LMIC contexts to conduct formal theoretical work; releasing all computational code and validation frameworks under open-source licenses; training early-career researchers in Nigeria through workshops on Bayesian hypothesis testing and computational pharmacology; and contributing to the development of addiction therapeutics that could reduce the global burden of substance use disorders, which disproportionately affect low- and middle-income countries. My qualifications for this work include sole-authored preprints on the CCT model, its formal mathematical specification, and its Bayesian population dynamics; a co-authored paper under review at Alcohol; a review article under review at Neuroscience and Biobehavioral Reviews; four open-source computational platforms; endorsements from leading computational and affective neuroscientists; and a provisional patent on the core architecture. I have no prior NSF funding and no current graduate enrollment, placing me squarely in the early-career, independent researcher category that this programme supports. BUDGET JUSTIFICATION The requested funds support eighteen months of independent research in Lagos, Nigeria, before anticipated MSc enrollment in October 2026. Total budget is 85,000 USD. Personnel: 45,000 USD. This covers a modest stipend for the PI (2,500 USD per month for 18 months) to allow full-time dedication to research without requiring concurrent employment. Nigerian cost of living is approximately one-third of U.S. levels, so this stipend is sufficient for housing, food, and local transportation in Lagos. Computing and Software: 15,000 USD. This covers cloud computing credits for Bayesian MCMC simulations on AWS/GCP (5,000 USD), a dedicated workstation with GPU for molecular dynamics simulations using GROMACS and AutoDock (8,000 USD), and software licenses for NEURON and Brian2 (2,000 USD). Current work is conducted on a personal laptop; a dedicated workstation will increase simulation throughput by approximately 10x. Publication and Dissemination: 10,000 USD. This covers open-access publication fees for two papers (approximately 3,000 USD each), travel to one international conference (4,000 USD for registration, airfare, and accommodation), and workshop materials for training sessions in Nigeria (1,000 USD). Data Management and Archiving: 5,000 USD. This covers DOI registration for all datasets and code repositories on Zenodo and OSF, long-term data storage, and maintenance of the IMPRINT, TOPOLOGIX, and GATE platforms on Supabase/Postgres. Indirect Costs: 10,000 USD. Calculated at 13.5 percent of direct costs, consistent with NSF policy for organizations without negotiated indirect cost rates. These funds support administrative coordination, including liaison with collaborators at Michigan, Harvard, Princeton, and NYU. Total: 85,000 USD. DATA MANAGEMENT PLAN All data generated during this project will be made publicly available under Creative Commons Attribution 4.0 International licenses. Three categories of data will be produced. First, simulation data from ODE/RK45 and Bayesian MCMC runs, including parameter estimates, posterior distributions, and convergence diagnostics. These will be deposited in Zenodo with DOIs and linked to the corresponding preprints. File formats will be CSV for tabular data and NetCDF for multidimensional arrays. Metadata will follow the Dublin Core standard. Second, source code for all simulations, including Python scripts using scipy, numpy, PyMC, and custom ODE solvers. Code will be hosted on GitHub under the Apache 2.0 license, with versioned releases archived on Zenodo. Documentation will include README files, example notebooks, and dependency specifications in conda environment files. Third, documentation of the discovery trajectory, including decision logs, failure records, and validation protocols. These will be published as supplementary materials to the meta-scientific case study and deposited in OSF. Data will be stored on institutional cloud storage (AWS S3 with versioning) during the project period, with daily backups. After project completion, data will be transferred to Zenodo and OSF for long-term preservation. No data will be subject to privacy or confidentiality restrictions, as no human subjects data will be collected during this phase of the project. CHECKLIST - [ ] Complete NSF Science of Science grant application on Grants.gov - [ ] Upload motivation letter (500 words maximum) - [ ] Upload research statement (600 words maximum) - [ ] Upload budget justification (300 words maximum) - [ ] Upload data management plan (300 words maximum) - [ ] Upload current CV with ORCID, GitHub, and publication links - [ ] Upload letters of endorsement from Kent Berridge, Samuel Gershman, Nathaniel Daw, and Marcelo Mattar - [ ] Upload provisional patent filing documentation (Q3 2026) - [ ] Upload preprints: foundational CCT paper, formal mathematical specification, Bayesian population dynamics - [ ] Upload proof of B.Pharm degree and PCN pharmacist license - [ ] Verify eligibility for early-career/independent researcher track - [ ] Confirm no current graduate enrollment status - [ ] Submit by programme deadline (check Grants.gov for exact date) EDITOR NOTES - Eligibility risk: NSF grants typically require U.S. institutional affiliation or eligibility determination for foreign organizations. Verify whether independent researchers based in Nigeria can apply directly or need a U.S.-based collaborator as PI. The profile lists no U.S. co-PI; this may need to be addressed. - Fact verification: Confirm that the provisional patent filing is indeed scheduled for Q3 2026 and that the patent office accepts filings from independent Nigerian inventors without legal representation. If not, budget may need to include patent attorney fees. - Gap: The profile does not specify whether Eniola has a U.S. bank account or institutional host for fund disbursement. NSF typically does not wire funds directly to individuals. A fiscal sponsor or U.S.-based collaborator may be required. - Gap: The profile lists no prior grant funding. The budget justification should explicitly state that this is the applicant's first grant application, which strengthens the early-career narrative but may require additional justification of financial management capacity. - Verification: Confirm that the three preprints are indeed sole-authored and that the ORCID and GitHub links are active and contain the described content. Reviewers will check these.