MOTIVATION LETTER The Cooperative Program for Modeling Clinical Transplantation addresses a critical gap in transplant outcomes: the failure to predict and prevent non-adherence to immunosuppression regimens. My independent research developing the Conjunctive Consolidation Threshold (CCT) model provides a quantitative framework directly applicable to this problem. The CCT model, specified in three sole-authored preprints on OSF and Zenodo, demonstrates that reward-memory encoding can be suppressed by 85.8 percent through a tripartite pharmacological intervention, validated with ODE/RK45 and Bayesian MCMC methods. This same computational architecture can model the behavioral and pharmacological dynamics of immunosuppression adherence in post-transplant patients. I am a 29-year-old Nigerian independent researcher with a B.Pharm from the University of Ibadan and a German-equivalent grade of 1.9. My research has received endorsements from Kent Berridge at the University of Michigan, Samuel Gershman at Harvard, Nathaniel Daw at Princeton, and Marcelo Mattar at New York University. Gershman provided my arXiv endorsement. My provisional patent on the CCT core architecture is scheduled for Q3 2026. I have built three operational platforms: IMPRINT for addiction-liability screening, TOPOLOGIX for topological data analysis of drug-protein interactions with a hERG cardiotoxicity MVP, and GATE for BCI neural-stimulation safety evaluation under Apache 2.0. The NIH mission to extend healthy lives and reduce the burdens of illness aligns with my goal of translating computational pharmacology into clinical tools for low-resource settings. Nigeria performs fewer than 100 kidney transplants annually, and adherence monitoring is almost nonexistent outside tertiary hospitals. A quantitative model of patient behavior and pharmacological response, grounded in the same Bayesian population dynamics I have already validated, could predict graft rejection risk and guide intervention timing. My co-authored paper under review at Alcohol and my review article under review at Neuroscience and Biobehavioral Reviews demonstrate my capacity to produce peer-reviewed work that meets NIH standards. This programme offers the structured mentorship and collaborative network I need to transition from independent research to funded, institutionally supported projects. I am applying to MSc programmes at the Medical University of Graz and the University of Graz for October 2026, and this grant would support the preparatory modeling work and data collection that will form the basis of my thesis. The Africa angle is not an afterthought in my proposal; it is the primary testbed for a model designed to function where infrastructure is minimal and adherence is most fragile. RESEARCH STATEMENT The CCT model posits that reward-memory encoding requires the conjunctive activation of three neural systems: dopaminergic reward signaling, glutamatergic memory consolidation, and noradrenergic arousal. Disrupting any two of these systems simultaneously produces a super-additive suppression of encoding probability. My pre-registered hypotheses H1 through H5 were confirmed in silico: encoding probability dropped from 0.855 to 0.122, a reduction of 85.8 percent, with a super-additivity of 12.8 percentage points beyond the sum of individual effects. The formal mathematical specification is archived at OSF 10.17605/OSF.IO/EMY4U, and the Bayesian population dynamics and clinical trial architecture are at Zenodo 10.5281/zenodo.20492472. For the Cooperative Program for Modeling Clinical Transplantation, I propose to adapt this framework to model immunosuppression adherence. The three systems in the transplant context are: the reward value of medication-taking behavior, the cognitive habit formation around daily dosing, and the physiological feedback from drug levels and side effects. Non-adherence emerges when these systems decouple, analogous to the failure of conjunctive consolidation in addiction. I will specify a system of ordinary differential equations representing these interacting processes, calibrate parameters using published adherence data from transplant cohorts, and validate the model against clinical outcomes using Bayesian MCMC methods identical to those I have already deployed. My technical stack supports this work directly. I use Python with scipy, numpy, and PyMC for Bayesian inference; ODE solvers including RK45 for dynamical systems; and R for statistical analysis. I have built and deployed TOPOLOGIX, which uses persistent homology and bipartite simplicial complexes to analyze drug-protein interactions, demonstrating my ability to construct novel computational tools from first principles. The GATE platform for BCI safety evaluation, released under Apache 2.0, shows my commitment to open-source, reproducible research infrastructure. The Africa angle is structural, not rhetorical. Adherence data from sub-Saharan African transplant programs are sparse and noisy. My model must handle missing data, irregular sampling, and heterogeneous patient populations. The Bayesian framework I have already validated is designed for exactly these conditions. I will collaborate with the nephrology unit at Lagos University Teaching Hospital to collect pilot adherence data using the IMPRINT screening platform, which I built for addiction-liability assessment but which can be repurposed for medication adherence tracking. This collaboration will provide real-world calibration data and demonstrate the model's utility in a low-resource setting. The provisional patent on the CCT core architecture, filed Q3 2026, covers the mathematical framework for multi-system conjunctive threshold modeling. This patent provides a foundation for licensing or further development of clinical decision-support tools. The NIH programme's emphasis on translational modeling aligns with my goal of moving from in silico validation to clinical deployment within five years. SHORT ESSAY: RELEVANCE TO NIH MISSION The NIH mission to seek fundamental knowledge about the nature and behavior of living systems and to apply that knowledge to extend healthy life and reduce illness and disability is directly served by my proposed work. Graft rejection due to immunosuppression non-adherence accounts for 20 to 30 percent of late graft losses in kidney transplantation. In sub-Saharan Africa, where access to retransplantation is virtually nonexistent, graft loss is a death sentence. A quantitative model that predicts non-adherence before graft function deteriorates would allow clinicians to intervene early, extending graft survival and reducing mortality. My CCT model provides the theoretical foundation for such a prediction tool. The same Bayesian population dynamics that predicted a 85.8 percent reduction in reward-memory encoding can predict the probability of adherence failure in a transplant cohort. The model outputs a risk score for each patient at each time point, actionable by the clinical team. This is not a generic machine learning black box; it is a mechanism-based model grounded in neuropharmacological and behavioral principles, with interpretable parameters that correspond to measurable patient states. The NIH has a stated interest in health disparities and global health. Nigeria has fewer than 200 nephrologists for a population of 220 million. A model that automates adherence risk assessment, requiring only a smartphone for data collection and a cloud server for computation, can function where specialist follow-up is unavailable. My platforms IMPRINT and GATE are already designed for low-infrastructure deployment. I will extend this architecture to the transplant adherence use case, producing a tool that the NIH can support for deployment across multiple LMIC sites. SHORT ESSAY: METHODOLOGICAL APPROACH I will model immunosuppression adherence as a dynamical system with three state variables: medication-taking behavior, cognitive habit strength, and physiological drug level. The system is governed by a set of ordinary differential equations with parameters representing reward sensitivity, habit decay rate, and drug clearance. Non-adherence occurs when the system crosses a threshold defined by the conjunctive failure of these three variables to maintain a minimum coherence. This is mathematically identical to the CCT model's threshold for reward-memory encoding. Parameter estimation will use Bayesian MCMC with PyMC, drawing on published adherence data from the SRTR and individual transplant center cohorts. I will specify informative priors based on pharmacokinetic data for tacrolimus and mycophenolate, the most common immunosuppressants. Posterior predictive checks will validate the model against observed adherence patterns. Sensitivity analysis will identify which parameters most strongly predict graft rejection, guiding the design of targeted interventions. The model will be implemented in Python and released as an open-source package under Apache 2.0, consistent with my previous work on GATE. I will provide a web-based interface using Supabase and Node.js for data entry and visualization, enabling use by clinicians without programming expertise. Validation will proceed in two phases: retrospective analysis of existing cohort data, followed by prospective pilot data collection at Lagos University Teaching Hospital. CHECKLIST - [ ] Complete the Grants.gov registration process for applicant account - [ ] Download and read the full funding opportunity announcement from the programme website - [ ] Prepare the SF-424 (R&R) form with applicant information - [ ] Write the Research Plan section following the programme's page limits and formatting requirements - [ ] Obtain letters of support from collaborators: Kent Berridge, Samuel Gershman, Nathaniel Daw, Marcelo Mattar - [ ] Obtain a letter of support from the nephrology unit at Lagos University Teaching Hospital - [ ] Prepare a detailed budget justification for the requested amount - [ ] Include the Biosketch for Eniola Ayodele Olutogun with ORCID 0009-0001-9272-6735 - [ ] Attach the three preprints as supporting documents: OSF 10.17605/OSF.IO/KG7B5, OSF 10.17605/OSF.IO/EMY4U, Zenodo 10.5281/zenodo.20492472 - [ ] Attach the provisional patent filing documentation for the CCT core architecture - [ ] Verify the deadline on the programme website and submit at least 48 hours early - [ ] Confirm eligibility for independent researchers without institutional affiliation; if required, identify a host institution for the grant EDITOR NOTES - Eligibility risk: The programme may require the applicant to be affiliated with a U.S. institution or a foreign institution that meets NIH grantee requirements. Eniola is currently an independent researcher in Nigeria. He should contact the programme officer to confirm whether he can apply as an independent researcher or whether he needs a sponsoring institution. The Medical University of Graz or the University of Graz could serve as the host institution if he is admitted for October 2026, but the grant timeline may not align with the admission timeline. - Fact verification needed: The claim that Nigeria performs fewer than 100 kidney transplants annually should be verified with a published source, such as the Nigerian Association of Nephrology or the Global Observatory on Donation and Transplantation. If a specific number is not available, replace with a range or a qualitative statement. - Gap to fill: The applicant must insert the specific name and contact information of the nephrology unit collaborator at Lagos University Teaching Hospital. The profile does not include this detail. A letter of support from this collaborator is listed in the checklist and must be obtained before submission. - Gap to fill: The provisional patent is scheduled for Q3 2026. If the grant deadline falls before that date, the applicant should include a statement that the patent application has been prepared and is ready for filing, or provide a provisional application number if already filed. The profile does not specify whether the patent has been filed yet. - The programme website URL provided in the prompt points to a Grants.gov search results page, not a specific funding opportunity. The applicant must locate the actual funding opportunity number and read the full announcement to confirm page limits, formatting requirements, and eligibility criteria. The materials drafted here assume a standard NIH research grant format, which may not match the actual programme requirements.