MOTIVATION LETTER A 29-year-old Nigerian pharmacist with no PhD, no lab, and no institutional backing has built a Bayesian-validated model of reward-memory encoding in addiction that confirmed all five pre-registered hypotheses. That model, the Conjunctive Consolidation Threshold, couples dopaminergic reward-prediction error, NMDAR-dependent long-term potentiation, and affective contrast into a single ODE system with 14 free parameters, calibrated via PyMC DEMetropolisZ against priors drawn from a systematic screen of 1,847 records. The posterior showed super-additivity of 13 to 22 percentage points across model versions. The work is sole-authored, pre-registered, and deposited on OSF and Zenodo. A co-authored paper is under review at Alcohol. Emergent Ventures funds people who produce real results without waiting for permission. That describes my entire research career. I have built a drug-resistance mutation predictor, TOPOLOGIX, that uses ESM-2 protein-language-model delta-embeddings and Morgan fingerprints with a Random Forest classifier. It achieves AUROC 0.804 on the Platinum benchmark, covering 100 percent of mutations versus roughly 18 percent for structure-limited tools like mCSM-lig, which scores around 0.70. I have built neurocascade, a receptor-to-behavior brain-circuit simulation engine that couples pharmacokinetics to Wilson-Cowan dynamics across three receptor systems, with 62 of 62 tests passing. I have built ergofluids, a Koopman-operator framework for modeling drug transport through tumor tissue, and reported its first real-data gate failure directly rather than reframing it. I have tested whether topological data analysis predicts hERG cardiotoxicity and found it does not beat a plain descriptor baseline, settling a question the literature had never actually run. These projects share a pattern: I identify a question, pre-register a protocol, execute the computation, and publish the result regardless of whether it confirms my hypothesis. That is the discipline I bring. I have endorsements from Kent Berridge at Michigan, Samuel Gershman at Harvard, Nathaniel Daw at Princeton, and Marcelo Mattar at NYU. None of them supervised me. They read my work. I am enrolled in the M.Sc. Digital Health at Hasso Plattner Institute starting winter 2026, but my research trajectory is independent and self-directed. Emergent Ventures funds the idea, not the institution. My idea is this: the CCT model can be extended from a theoretical framework into a clinically actionable screening tool for addiction liability, and TOPOLOGIX can be deployed as a free, open-access service for researchers in low-resource settings who cannot afford structure-based prediction. Nigeria has one of the highest rates of substance use disorder in West Africa and one of the lowest rates of computational pharmacology research. I am positioned at that intersection. I am applying for runway to execute the next set of pre-registered experiments. SHORT ESSAY: RESEARCH VISION My research vision has two tracks that converge on a single goal: making computational pharmacology predictive and accessible from any laboratory in the world, regardless of infrastructure. The first track is addiction neuroscience. The CCT model formalizes a tripartite threshold for reward-memory encoding: dopamine reward-prediction error, NMDAR-dependent plasticity, and affective contrast must all cross a conjunctive threshold for a memory to consolidate. The model is mathematically specified, Bayesian-calibrated, and confirmed against five pre-registered hypotheses. The next step is to fit it to real behavioral data from rodent self-administration paradigms, which requires access to time-series intracranial recordings and lever-press logs. I have the computational framework. I need the collaboration and the compute to calibrate the circuit-layer parameters that are currently labeled illustrative. The second track is drug-resistance prediction. TOPOLOGIX already beats structure-based tools on coverage and matches them on accuracy. The limitation is that it was trained on mutation data from the Platinum benchmark, which is small and biased toward well-studied proteins. I want to expand the training set by generating synthetic mutation data via directed evolution simulations in silico, then retrain the ESM-2 delta-embedding pipeline on that augmented corpus. If the AUROC holds above 0.80 on held-out families, the tool becomes a general-purpose resistance predictor that requires only a protein sequence and a drug SMILES string. These two tracks share a methodological core: pre-registered, Bayesian-calibrated computational models that make falsifiable predictions. I do not build tools for their own sake. I build them to answer specific questions about how molecules and circuits behave, and I test those answers against data. SHORT ESSAY: WHY EMERGENT VENTURES Emergent Ventures selects for three things: novelty of ideas, evidence of independent execution, and potential for high impact. My work scores on all three. The CCT model is novel. No existing pharmacological framework for addiction prevention couples reward-prediction error, LTP, and affective contrast into a single ODE threshold. The model is 14 equations with Bayesian posteriors. That is the kind of specificity Emergent Ventures rewards. The evidence of independent execution is documented across five pre-registered projects, three sole-authored preprints, one co-authored paper under review, and a public GitHub repository with working code for every model. I have no PhD supervisor, no lab head, no grant from a funding agency. Every result I have produced came from a personal laptop, open-source libraries, and literature-elicited priors. The potential for high impact is concrete. Addiction is a global health crisis with few computational tools for prevention. Drug resistance is a growing threat with prediction tools that cover less than 20 percent of known mutations. My models address both problems directly, and they are designed to be deployed without expensive infrastructure. A researcher in Lagos with a laptop and an internet connection can run TOPOLOGIX on a new mutation today. That is the kind of accessibility that changes who gets to do science. Emergent Ventures also funds people who are early in their trajectory and unlikely to fit traditional grant categories. I am 29, Nigerian, pre-PhD, and independent. I do not fit the NIH early-career mold or the ERC Starting Grant mold. I fit the Emergent Ventures mold. EDITOR NOTES - Eligibility risk: Emergent Ventures typically awards between $5,000 and $50,000. The applicant's stated floor is $10,000, which is within range, but larger requests may require stronger justification. The letter does not specify an amount. The applicant should decide on a specific figure before submitting. - Fact verification needed: The applicant's profile states "co-authored paper in Alcohol (Elsevier, under review)." The status of that paper should be confirmed before submission. If it has been accepted or rejected, the letter should reflect the current status. - Personal detail gap: The applicant's profile does not include a specific dollar amount requested, a timeline for the proposed work, or a budget breakdown. Emergent Ventures applications typically benefit from a clear, concise budget. The applicant should prepare a one-paragraph budget specifying what the funds will cover (compute time, data access, conference travel, etc.). - Endorsement context: The letter mentions endorsements from Berridge, Gershman, Daw, and Mattar. The applicant should confirm that these individuals are willing to be named as endorsers and, if possible, provide brief letters of support or email confirmations. - Programme URL note: The URL provided in the prompt points to a blog post, not the official Emergent Ventures application page. The applicant should verify the correct application portal and submission instructions before submitting.