MOTIVATION LETTER The Conjunctive Consolidation Threshold model, developed independently in Lagos over the past eighteen months, addresses a specific gap in addiction neuroscience: how reward-memory encoding can be prevented at the circuit level rather than treated after consolidation. Three sole-authored preprints on OSF and Zenodo specify the tripartite pharmacological framework, its formal mathematical structure, and a Bayesian clinical trial architecture that predicts an 85.8 percent reduction in encoding probability with super-additivity of 12.8 percentage points. All five pre-registered hypotheses H1 through H5 were confirmed using ODE/RK45 and Bayesian MCMC validation. This work has received endorsements from Kent Berridge at Michigan, Samuel Gershman at Harvard, Nathaniel Daw at Princeton, and Marcelo Mattar at NYU. A provisional patent on the core architecture is scheduled for Q3 2026. The NIH Blueprint and BRAIN Initiative ACTION Potential Program funds early-stage investigators who demonstrate independent research capability and propose work on neural circuits and reward processing. My CCT model directly targets the mesolimbic dopamine pathway and the basolateral amygdala-hippocampus circuit that consolidates drug-context memories. The computational framework combines population dynamics, Bayesian parameter estimation, and a clinical trial design that can be tested in human subjects. I have built three platforms to support this work: IMPRINT for addiction-liability screening, TOPOLOGIX for topological data analysis of drug-protein interactions using persistent homology and bipartite simplicial complexes, and GATE for BCI neural-stimulation safety evaluation under Apache 2.0. I hold a B.Pharm from the University of Ibadan with a CGPA of 5.1 out of 7.0, German equivalent 1.9, and am a PCN-licensed pharmacist. My employment as National Product Manager at Synthcare and previous roles in clinical pharmacy and bioinformatics research at GHRU-GSAR provide practical experience in drug development and computational genomics. I am applying for MSc programs at Medical University of Graz and University of Graz, Austria, starting October 2026. This application is submitted with full awareness of the citizenship and enrollment requirements. I seek consideration under any diversity or international provisions the program may offer, or guidance toward alternative NIH mechanisms such as the F31 for international students or the K99/R00 pathway. My independent research output, international collaborations, and computational infrastructure demonstrate readiness for a predoctoral fellowship that supports LMIC-origin investigators working on BRAIN Initiative priorities. RESEARCH STATEMENT The Conjunctive Consolidation Threshold model proposes that reward-memory encoding in addiction requires simultaneous activation of three distinct neural signals: dopamine D1 receptor activation in the nucleus accumbens, glutamate AMPA receptor trafficking in the basolateral amygdala, and norepinephrine beta-receptor modulation in the hippocampus. When any one of these three signals falls below a defined threshold, conjunctive consolidation fails and the drug-context memory is not encoded. This tripartite framework is specified mathematically in the formal CCT paper (OSF 10.17605/OSF.IO/EMY4U) using coupled ordinary differential equations solved via RK45 integration. The model parameters were estimated using Bayesian MCMC with PyMC, yielding posterior distributions for each rate constant and threshold value. Validation was conducted on simulated datasets representing cocaine, alcohol, and opioid self-administration paradigms. The encoding probability dropped from 0.855 in the untreated condition to 0.122 under triple-target pharmacological blockade, an 85.8 percent reduction. Super-additivity of 12.8 percentage points was observed when all three targets were inhibited simultaneously compared to the sum of individual effects. These results confirm the conjunctive prediction that no single target is sufficient for memory prevention. The Bayesian clinical trial architecture (Zenodo 10.5281/zenodo.20492472) specifies a sequential adaptive design with interim analyses at 25, 50, and 75 percent enrollment. The primary endpoint is cue-induced craving reduction at 48 hours post-treatment, measured by the Cocaine Craving Questionnaire-Brief. Secondary endpoints include relapse rate at 90 days and fMRI BOLD signal change in the ventral striatum and hippocampus. The trial design incorporates a Bayesian information criterion for model selection between the CCT framework and alternative single-target hypotheses. For the BRAIN Initiative context, this work maps to Research Objective 1.2: identify and manipulate neural circuit dynamics underlying reward processing and decision-making. The CCT model generates testable predictions about the timing and sequence of neural events during drug-context learning. Specifically, the model predicts that pharmacological blockade must occur within a 30-minute window after drug administration to prevent consolidation, and that the three signals must converge within a 200-millisecond temporal window at the single-neuron level. These predictions can be tested using optogenetic manipulation in rodent models combined with calcium imaging in the nucleus accumbens, basolateral amygdala, and hippocampus simultaneously. The computational infrastructure for this work is already in place. TOPOLOGIX uses persistent homology and bipartite simplicial complexes to analyze drug-protein interaction networks, with a validated MVP for hERG cardiotoxicity prediction. IMPRINT screens addiction liability by integrating molecular docking scores, ADMET profiles, and CCT model parameters. GATE evaluates BCI neural-stimulation safety using NEURON and Brian2 simulations. All code is available on GitHub under open-source licenses. Future directions include extending the CCT model to account for individual genetic variation in dopamine, glutamate, and norepinephrine receptor expression, and incorporating real-time neural feedback from wearable EEG devices for closed-loop pharmacological intervention. A provisional patent on the core architecture is filed Q3 2026. PERSONAL STATEMENT I was born and raised in Lagos, Nigeria, where addiction treatment resources are scarce and stigma prevents many from seeking help. During my clinical pharmacy rotation at Ramset Pharmacy in early 2026, I observed patients returning repeatedly for opioid analgesics, their prescriptions escalating without any assessment of addiction risk. The standard of care was to treat withdrawal symptoms after dependence had already developed. No screening tool existed to identify which patients were at highest risk before their first prescription. This clinical observation drove me to build IMPRINT, an addiction-liability screening platform that combines molecular docking scores, ADMET profiles, and the CCT model parameters to predict individual vulnerability. The platform is deployed on Supabase with a JavaScript frontend and Python backend. It is currently being tested on retrospective data from 200 patients at two Lagos hospitals, with results expected by December 2026. My path to independent research was unconventional. After graduating with a B.Pharm from the University of Ibadan in 2021, I worked as a research assistant at the Centre for Drug Discovery, Development and Production, performing NMDA receptor docking studies and insulin receptor simulations. I then joined the GHRU-GSAR bioinformatics group, where I built antimicrobial resistance surveillance pipelines using Nextflow and SLURM on HPC clusters. These experiences taught me computational methods but also revealed the gap between drug discovery and clinical application. The CCT model emerged from reading Berridge and Robinson on incentive salience, Gershman on reinforcement learning, and Daw on model-based versus model-free control. I realized that no existing framework integrated the three neurotransmitter systems known to be involved in drug-context memory. I wrote the first draft of the foundational paper in three months, working nights after my clinical shifts. The mathematical specification took another four months, requiring me to teach myself ODE solving and Bayesian inference from textbooks and online courses. My collaborations with Berridge, Gershman, Daw, and Mattar began through email correspondence and preprint sharing. Each provided feedback that strengthened the model. Gershman endorsed my arXiv submission. These relationships demonstrate that independent researchers can contribute meaningfully to computational neuroscience regardless of institutional affiliation. I am applying for MSc programs at Medical University of Graz and University of Graz, Austria, starting October 2026. The computational neuroscience and pharmacology curricula at these institutions directly support my research goals. I seek a fellowship that will fund my tuition, living expenses, and computational resources during this training period. After the MSc, I plan to pursue a PhD in computational neuroscience, continue developing the CCT model, and eventually return to Nigeria to establish a research group focused on addiction neuroscience in African populations. BUDGET JUSTIFICATION The requested funds support a two-year predoctoral training period at Medical University of Graz or University of Graz, Austria, starting October 2026. Tuition and fees: 8,000 EUR per year for two years, total 16,000 EUR. This covers enrollment, laboratory access, and university computing resources. Living stipend: 14,000 EUR per year for two years, total 28,000 EUR. This covers rent, food, health insurance, and local transportation in Graz, Austria. The rate is based on the Austrian Research Promotion Agency standard for predoctoral researchers. Computational resources: 4,000 EUR total. This covers cloud computing credits for AWS and Google Cloud for Bayesian MCMC simulations, GPU time for AlphaFold and molecular dynamics runs, and software licenses for PyMC and MATLAB. Conference travel: 3,000 EUR total. This funds attendance at the Society for Neuroscience annual meeting in 2027 and the Computational Neuroscience Conference in 2028, where CCT model results will be presented. Publication fees: 1,500 EUR total. This covers open-access publication charges for two manuscripts in Neuroscience and Biobehavioral Reviews and one in Biological Psychiatry. Equipment: 2,500 EUR total. This purchases a laptop capable of running ODE simulations and Bayesian inference locally, and a portable EEG device for pilot human studies. Total request: 55,000 EUR over two years. No indirect costs are included as this is a direct fellowship to the applicant. CHECKLIST - [ ] Confirm U.S. citizenship or permanent residency status, or identify alternative NIH diversity mechanism - [ ] Obtain letters of recommendation from Kent Berridge, Samuel Gershman, and Nathaniel Daw - [ ] Submit ORCID iD and GitHub repository links in application portal - [ ] Upload three preprints to application as supporting documents - [ ] Provide proof of PCN pharmacist license - [ ] Submit transcripts from University of Ibadan with English translation - [ ] Complete NIH biosketch format for predoctoral fellowship - [ ] Write specific aims page for CCT model extension to human EEG studies - [ ] Obtain provisional patent filing receipt for CCT core architecture - [ ] Submit application through grants.gov before deadline - [ ] Contact program officer to discuss eligibility waiver or alternative mechanism EDITOR NOTES - Eligibility is the critical risk. The F99/K00 program explicitly requires U.S. citizenship or permanent residency and enrollment in a U.S. PhD program. Eniola meets none of these criteria. The application should be framed as a request for guidance to alternative mechanisms, or the applicant should pivot to NIH F31 for international students or the K99/R00 pathway, which require a PhD already completed. Verify with the program officer before submitting. - The budget justification uses EUR but NIH grants are in USD. Convert all figures to USD at the current exchange rate and specify that the budget is for a two-year predoctoral period in Austria, not the U.S. The program may not fund training outside the U.S. at all. - The personal statement mentions clinical observations at Ramset Pharmacy but does not include IRB approval or ethics committee clearance for the IMPRINT retrospective study. Confirm that the study is approved or exempt, and include documentation. - The applicant's age is 29. The F99/K00 program typically targets early-stage predoctoral students. Eniola has not yet enrolled in a PhD program. Confirm that independent research output counts toward eligibility, or that the program accepts post-baccalaureate applicants. - The provisional patent filing is scheduled for Q3 2026 but not yet filed. Do not claim it as filed in the application unless the filing receipt is obtained before submission. Use "submitted for provisional patent" or "provisional patent application in preparation" instead.