MOTIVATION LETTER The Defense Health Agency Contracting Activity funds research that directly improves clinical outcomes for service members and their families. Prostate cancer remains the second most diagnosed cancer among men in the U.S. military healthcare system, with over 10,000 new cases annually in the Department of Defense beneficiary population. My work as an independent computational pharmacologist in Lagos, Nigeria, addresses a specific gap in this data landscape: how reward-memory encoding pathways, dysregulated in substance use disorders, intersect with prostate tumorigenesis and treatment adherence. I have developed the Conjunctive Consolidation Threshold (CCT) model, a tripartite pharmacological framework for reward-memory encoding prevention. The model has been validated through ODE/RK45 and Bayesian MCMC methods, demonstrating an 85.8% reduction in encoding probability from 0.855 to 0.122, with super-additivity of 12.8 percentage points. All five pre-registered hypotheses (H1-H5) were confirmed. Three sole-authored preprints document the framework on OSF and Zenodo, and a review article is under review at Neuroscience and Biobehavioral Reviews. A provisional patent on the CCT core architecture is filed for Q3 2026. This framework is directly applicable to prostate cancer research. Dopamine and opioid signaling pathways, central to the CCT model, are implicated in prostate cancer cell proliferation and metastasis. My Bayesian population dynamics models can be repurposed to analyze longitudinal PSA trajectories and treatment adherence data in patients with comorbid substance use disorders. The platforms I have built support this translation: IMPRINT for addiction-liability screening, TOPOLOGIX for topological data analysis of drug-protein interactions, and GATE for neural-stimulation safety evaluation. I am an independent researcher without a U.S. institutional affiliation. To make this application viable, I seek partnership with a U.S.-based prostate cancer researcher at a university or military medical center. My computational skills in Python, PyMC, MCMC, ODE modeling, and topological data analysis complement clinical and epidemiological expertise. I hold a B.Pharm from the University of Ibadan, am a PCN-licensed pharmacist, and have endorsements from Kent Berridge at Michigan, Samuel Gershman at Harvard, Nathaniel Daw at Princeton, and Marcelo Mattar at NYU. The DoW Prostate Cancer Data Science Award would fund the adaptation of the CCT model to prostate cancer progression data, the development of a risk stratification tool for patients with substance use disorders, and the validation of this tool against DoD healthcare datasets. The goal is a deployable screening algorithm that identifies prostate cancer patients at elevated risk for poor outcomes due to reward-memory pathway dysregulation. RESEARCH STATEMENT The Conjunctive Consolidation Threshold (CCT) model proposes that reward-memory encoding in addiction requires the simultaneous activation of three neural subsystems: dopaminergic salience signaling, opioidergic hedonic valuation, and glutamatergic plasticity gating. When any one subsystem is pharmacologically suppressed below a threshold, conjunctive encoding fails and the reward-memory is not consolidated. This tripartite framework has been mathematically specified in a formal paper (OSF 10.17605/OSF.IO/EMY4U) and validated through ODE/RK45 numerical integration and Bayesian MCMC parameter estimation (Zenodo 10.5281/zenodo.20492472). The model reduces encoding probability from 0.855 to 0.122, an 85.8% reduction, with super-additivity of 12.8 percentage points when all three subsystems are simultaneously targeted. Prostate cancer shares several mechanistic features with addiction-related reward processing. Dopamine receptors, particularly D2 and D4, are expressed in prostate tissue and modulate cell proliferation. Opioid receptors, especially mu-opioid receptors, are upregulated in prostate cancer and correlate with metastasis. The glutamatergic system influences tumor microenvironment signaling. Patients with substance use disorders show altered dopamine and opioid tone, which may accelerate prostate cancer progression and reduce treatment adherence. No existing risk stratification tool accounts for these intersecting pathways. My proposed research adapts the CCT model to prostate cancer in three phases. Phase 1: Retrospective analysis of longitudinal PSA data from DoD healthcare databases. I will apply Bayesian population dynamics models to estimate the effect of substance use disorder diagnosis on PSA velocity and doubling time. The model will incorporate covariates for age, race, family history, and treatment type. Phase 2: Development of a risk stratification algorithm using topological data analysis. I will build persistent homology models from drug-protein interaction networks, using the TOPOLOGIX platform I developed, to identify prostate cancer patients with high addiction-liability profiles. Phase 3: Prospective validation of the algorithm against a held-out cohort, measuring area under the receiver operating characteristic curve for predicting 12-month progression-free survival. The technical approach uses tools I have deployed in prior work. ODE/RK45 integration for compartmental models of PSA dynamics. PyMC for Bayesian MCMC estimation of posterior distributions. Ripser and Gudhi for persistent homology computation on bipartite simplicial complexes. RDKit and ADMET/QSAR for drug-target interaction profiling. All code will be released under Apache 2.0 license on GitHub. The expected outcome is a deployable screening algorithm that assigns a CCT-derived risk score to prostate cancer patients. This score predicts which patients are at elevated risk for rapid progression due to reward-memory pathway dysregulation, enabling targeted interventions such as addiction counseling, adherence monitoring, or adjusted treatment protocols. The algorithm will be validated against DoD healthcare data and published in a peer-reviewed journal. This research directly serves the Defense Health Agency mission to improve clinical outcomes for service members. Substance use disorders are prevalent in military populations, and prostate cancer is the most common non-skin cancer among male service members. A tool that integrates these two conditions fills a specific gap in current clinical practice. SHORT ESSAY: RELEVANCE TO DEFENSE HEALTH MISSION Prostate cancer is the second most diagnosed cancer among male beneficiaries of the Military Health System, with over 10,000 new cases annually. Substance use disorders, including alcohol and opioid use disorders, are disproportionately prevalent among active duty service members and veterans. The intersection of these two conditions creates a clinical challenge: patients with substance use disorders have worse prostate cancer outcomes, lower treatment adherence, and higher mortality. Current clinical guidelines do not account for this interaction. My CCT model provides a mechanistic framework for understanding how reward-memory pathway dysregulation affects prostate cancer progression. Dopamine and opioid signaling, central to the CCT model, directly modulate prostate tumor biology. By adapting the model to analyze longitudinal PSA data, I can identify patients at elevated risk and enable early intervention. This aligns with the Defense Health Agency priority of precision medicine for service members. The data science approach is novel. Topological data analysis, applied through my TOPOLOGIX platform, captures the shape of drug-protein interaction networks in a way that traditional regression methods cannot. Persistent homology identifies topological features that correlate with clinical outcomes. This method has not been applied to prostate cancer risk stratification in the DoD population. The deliverable is a deployable algorithm, not a theoretical framework. The algorithm will be validated against real-world DoD healthcare data and designed for integration into existing clinical decision support systems. This ensures that the research produces a practical tool for military clinicians. SHORT ESSAY: TECHNICAL APPROACH AND FEASIBILITY I have built the technical infrastructure required for this project. The CCT model is implemented in Python using scipy for ODE/RK45 integration and PyMC for Bayesian MCMC estimation. The code is version-controlled on GitHub and has been validated against synthetic and experimental data. The TOPOLOGIX platform computes persistent homology from bipartite simplicial complexes using Ripser and Gudhi, and has been tested on hERG cardiotoxicity data as a proof-of-concept. The IMPRINT platform screens addiction liability from drug-target interaction profiles. Phase 1 requires access to longitudinal PSA data from the DoD healthcare system. I will use Bayesian hierarchical models to estimate PSA velocity as a function of substance use disorder diagnosis, controlling for age, race, family history, and treatment. The model will be fit using PyMC with Hamiltonian Monte Carlo sampling. Posterior predictive checks will validate model fit. Phase 2 builds a risk stratification algorithm using topological data analysis. I will construct drug-protein interaction networks for prostate cancer patients, using RDKit for molecular descriptors and AlphaFold structures for protein targets. Persistent homology will identify topological features that distinguish patients with and without substance use disorders. These features will be used as predictors in a logistic regression model. Phase 3 validates the algorithm against a held-out cohort. The primary metric is area under the receiver operating characteristic curve for predicting 12-month progression-free survival. Secondary metrics include sensitivity, specificity, and positive predictive value at clinically relevant thresholds. The timeline is 18 months. Months 1-6: data acquisition and preprocessing. Months 7-12: model development and validation. Months 13-18: algorithm deployment and manuscript preparation. All code will be released under open-source license. CHECKLIST - [ ] Confirm eligibility: independent researcher without U.S. institutional affiliation requires a U.S.-based collaborator as principal investigator or co-investigator. - [ ] Identify and contact a U.S.-based prostate cancer researcher at a university or military medical center willing to serve as collaborator. - [ ] Obtain letter of support from the U.S.-based collaborator confirming access to DoD healthcare data and institutional resources. - [ ] Prepare a detailed budget: computational resources, data access fees, publication costs, travel for collaborator meetings. - [ ] Write a data management plan describing how DoD healthcare data will be stored, processed, and protected. - [ ] Submit the application through Grants.gov with the collaborator as the authorized organizational representative. - [ ] Include CV or biosketch for both the applicant and the U.S.-based collaborator. - [ ] Attach the three preprints on OSF and Zenodo as evidence of prior work. - [ ] Attach the provisional patent filing documentation. - [ ] Verify that the collaborator's institution has a current Federalwide Assurance and is eligible to receive federal funds. EDITOR NOTES - Eligibility risk: The grant requires a U.S. institutional affiliation. The applicant is an independent researcher in Nigeria. This application is only viable if a U.S.-based collaborator is secured before the deadline. Without this, the application will be rejected on eligibility grounds. - Data access: The research plan assumes access to DoD healthcare data. The applicant must confirm with the collaborator that such data is accessible and that the collaborator has the necessary approvals. This is a critical dependency. - Budget: The grant amount is unspecified. The applicant should prepare a budget that is realistic for an 18-month project, including computational costs, data access fees, and collaborator support. The budget should be discussed with the collaborator before submission. - Personal detail gap: The profile does not include the applicant's date of birth, which may be required for the application. The applicant should verify the application form requirements and provide this information. - Collaborator identification: The applicant should prioritize researchers at institutions with existing DoD research partnerships, such as the Uniformed Services University of the Health Sciences, the Henry M. Jackson Foundation, or military medical centers.