MOTIVATION LETTER The quantitative methods I have spent four years building as an independent computational researcher are directly transferable to the impact evaluation problems at the heart of ISSER's mandate. My work has centered on Bayesian calibration of complex dynamical systems, specifically the CCT model, a tripartite pharmacological framework for reward-memory encoding in addiction. That model required fitting 14 free parameters to literature-elicited priors drawn from a systematic screen of 1,847 records, using PyMC's DEMetropolisZ sampler, and confirming all five pre-registered hypotheses with posterior super-additivity of 13 to 22 percentage points across model versions. The same methodological discipline, the same commitment to pre-registration and honest reporting of negative results, applies directly to agricultural systems where intervention effects propagate through nonlinear feedback loops. My training as a pharmacist at the University of Ibadan, where I graduated with a CGPA of 5.1 out of 7.0, gave me a grounding in empirical research and a lived understanding of food systems and rural health in Nigeria. I have since extended that foundation through independent computational research, including a pre-registered replication study on hERG cardiotoxicity where I discovered that topological descriptors do not beat a plain baseline (AUROC 0.8426 versus 0.8782), a result I reported directly rather than reframing. That experience taught me that rigorous null results are as valuable as positive findings, a principle I would bring to impact evaluation at ISSER. ISSER's selection criteria emphasize strong quantitative skills, experience in data collection, and demonstrated interest in agriculture and food systems in developing countries. My current work on TOPOLOGIX, which predicts drug-resistance mutations from protein language model embeddings with an AUROC of 0.804 on the Platinum benchmark, demonstrates my capacity to build predictive models from complex biological data. The methodological core, feature engineering, model validation, and honest benchmarking, is identical to what impact evaluation requires when estimating treatment effects from observational or quasi-experimental data. I am currently enrolled in the M.Sc. Digital Health program at the Hasso Plattner Institute in Potsdam, Germany, starting Winter Semester 2026/27. This fellowship represents a strategic step toward doctoral studies in development economics or agricultural economics, where I intend to apply systems-dynamics and machine-learning methods to questions of agricultural intervention design and food security in West Africa. My Nigerian background and my experience working as a clinical pharmacist in Lagos give me both the motivation and the contextual knowledge to contribute meaningfully to ISSER's research agenda. RESEARCH STATEMENT The central question guiding my proposed research at ISSER is whether systems-dynamics models, calibrated with Bayesian methods, can improve the design and evaluation of agricultural interventions in Sub-Saharan Africa. Standard impact evaluation methods, including randomized controlled trials and difference-in-differences designs, estimate average treatment effects at specific points in time. They struggle, however, to capture how intervention effects propagate through coupled systems, such as the interaction between fertilizer adoption, soil degradation, credit access, and household nutrition. My proposed research addresses this gap by developing a framework that combines the rigor of causal identification with the realism of dynamical systems. My methodological foundation comes from four years of independent computational research. The CCT model, my primary line of work, is a tripartite ODE framework for reward-memory encoding that couples dopaminergic reward prediction error, NMDAR-dependent long-term potentiation, and affective contrast into a single system. I calibrated this model using Bayesian MCMC with 14 free parameters, eliciting priors from a systematic screen of 1,847 records, and confirmed all five pre-registered hypotheses. The model produced posterior super-additivity of 13 to 22 percentage points across versions, demonstrating that coupled nonlinear systems can capture dynamics that single-equation models miss. The same architecture, a system of coupled differential equations with literature-informed priors and Bayesian calibration, applies directly to agricultural systems where soil nutrients, water availability, labor allocation, and market prices form a coupled dynamical system. My experience with honest negative results is equally relevant. In my cardiotoxicity topology study, I tested whether bipartite persistent homology predicts hERG cardiotoxicity from protein-ligand interface geometry. A pre-registered, powered replication found that topological features do not beat a plain descriptor baseline, AUROC 0.8426 versus 0.8782. I reported this null result directly. In my interface-topology-for-resistance study, the same class of topological constructs carried almost no signal for drug-resistance prediction, AUROC 0.425 and 0.485 on the Platinum benchmark. These results taught me that pre-registration and honest reporting are not bureaucratic constraints but scientific tools. Impact evaluation in agriculture faces the same publication bias pressures, and I would bring this discipline to ISSER's research program. My current work on TOPOLOGIX demonstrates my capacity to build predictive models from complex data. Using ESM-2 protein language model delta-embeddings combined with Morgan fingerprints and a Random Forest classifier, I achieved an AUROC of 0.804 plus or minus 0.025 on the Platinum benchmark of 553 mutations, outperforming structure-based baselines such as mCSM-lig at approximately 0.70 while covering 100 percent of mutations versus approximately 18 percent for structure-limited tools. The methodological lesson, that sequence-based representations can outperform structure-based ones when coverage is limited, transfers directly to agricultural settings where detailed structural data on smallholder farms is scarce but observational data is abundant. I propose a two-paper research agenda for the fellowship. The first paper will develop a Bayesian-calibrated systems-dynamics model of a specific agricultural intervention, such as fertilizer subsidy programs or irrigation adoption, using data from a West African setting. The model will be pre-registered, with priors elicited from published agricultural economics literature, and validated against held-out data. The second paper will compare the predictive performance of this systems-dynamics approach against standard econometric methods, quantifying the conditions under which dynamical modeling adds value over reduced-form estimation. Both papers will follow the pre-registration and honest-reporting protocols I have used in my computational research. This agenda aligns directly with ISSER's focus on agriculture, natural resources, and food systems in developing countries. It leverages my existing quantitative skills, including Python, PyMC, ODE solvers, and machine learning, while building the econometric toolkit I would need for doctoral studies. The fellowship would provide the institutional structure and mentorship necessary to translate my computational methods into development research, and I am prepared to produce at least two technical papers during the fellowship period. SHORT ESSAY: MOTIVATION FOR DOCTORAL STUDIES My path to doctoral studies in development economics runs through computational modeling. I began as a pharmacist at the University of Ibadan, where I graduated with a CGPA of 5.1 out of 7.0, and spent two years as a clinical pharmacist in Lagos before moving into computational research. That transition was driven by a conviction that complex systems, whether pharmacological or agricultural, cannot be understood through single-variable analysis alone. My independent research since 2024 has tested that conviction across multiple domains, from addiction neuroscience to drug-resistance prediction, and has produced both confirmations and honest null results. The ISSER Pre-Doctoral Fellowship is the right next step because it offers structured mentorship in impact evaluation, a methodological toolkit I have not yet formally trained in, while allowing me to apply the quantitative skills I have already developed. My Bayesian calibration experience, my pre-registration discipline, and my software engineering capabilities are directly transferable to agricultural systems research. What I lack is formal training in econometrics and exposure to the development economics literature, gaps this fellowship would fill. My long-term goal is to establish a computational development economics research group in Nigeria, focused on agricultural systems modeling and food security. The fellowship would provide the technical foundation, the publication record, and the mentorship network necessary to pursue doctoral studies and eventually return to West Africa to build that group. I am committed to producing rigorous, pre-registered research that meets international standards while addressing questions of direct relevance to African agricultural policy. CHECKLIST - [ ] Verify ISSER Pre-Doctoral Fellowship 2026-2027 eligibility requirements on the official ISSER website, not the AfterSchoolAfrica aggregator page - [ ] Confirm whether a Master's degree in economics, agricultural economics, or a related field is strictly required or whether the M.Sc. Digital Health enrollment satisfies the requirement - [ ] Obtain official transcripts from University of Ibadan and Hasso Plattner Institute, including German equivalent grade documentation - [ ] Request letters of recommendation from at least two of the following: Kent Berridge (Michigan), Samuel Gershman (Harvard), Nathaniel Daw (Princeton), Marcelo Mattar (NYU) - [ ] Prepare a CV formatted to ISSER specifications, emphasizing quantitative methods and African research experience - [ ] Draft a two-page research proposal expanding the systems-dynamics impact evaluation agenda described in the research statement - [ ] Gather documentation of pre-registrations for the CCT model, hERG topology study, and TOPOLOGIX benchmark results - [ ] Prepare a writing sample, likely the CCT model preprint or the hERG topology replication paper - [ ] Confirm the application deadline and submission portal directly with ISSER - [ ] Prepare a statement on how the fellowship fits with the M.Sc. Digital Health program timeline at HPI/Potsdam EDITOR NOTES - Eligibility risk: The ISSER selection criteria list a Master's degree in economics, agricultural economics, or a related field as required. The applicant holds a B.Pharm and is enrolled in an M.Sc. Digital Health program that has not yet started. A human reviewer must verify whether ISSER will accept this combination or whether the fellowship requires a completed Master's degree in a social science field. If this is a hard requirement, the application should not proceed without clarification. - The framing angle pivots the applicant's computational modeling experience toward agricultural systems and impact evaluation. This is a legitimate transfer of methods, but the applicant has no formal training or publication record in agricultural economics. The research statement should be reviewed by someone familiar with ISSER's current SPIA research agenda to confirm the proposed systems-dynamics approach is actually relevant to their ongoing projects, not just plausible in the abstract. - The applicant must insert specific details about which West African agricultural dataset or intervention they would study. The current draft names fertilizer subsidies and irrigation adoption as examples but does not commit to a specific dataset, country, or policy question. ISSER reviewers will expect a concrete research plan, not illustrative examples. - The motivation letter and research statement reference the CCT model and TOPOLOGIX extensively. These are biomedical projects. The applicant should be prepared to explain, in interview or in a supplementary document, why a fellowship focused on African development should fund someone whose primary research output has been in addiction neuroscience and protein machine learning. The transferable-skills argument is strong but needs to be made explicitly and repeatedly. - The applicant's employment history includes a role as National Product Manager at Synthcare starting March 2026. This is a commercial role. ISSER may view this as a conflict with full-time fellowship research. The applicant should clarify whether the fellowship would be part-time or whether they would need to pause or resign from this position.