MOTIVATION LETTER Antimicrobial resistance is projected to cause ten million deaths per year by 2050, with sub-Saharan Africa carrying a disproportionate share of that burden. In Nigeria, where I am licensed as a pharmacist and where multidrug-resistant tuberculosis and resistant Gram-negative infections are routine clinical realities, the gap between genomic sequence data and actionable treatment decisions is measured in months and years, not days. My research program attacks that gap directly. I am applying to the OWSD PhD Fellowship because it is the only programme I have found that combines three things I need: funding for computational research in a low-resource setting, a mandate to build scientific capacity in Science and Technology Lagging Countries, and a requirement that fellows return home to strengthen local institutions. That return is the entire point of my work. I hold a B.Pharm from the University of Ibadan with a German-equivalent grade of 1.9, and I am currently enrolled in the M.Sc. Digital Health programme at the Hasso Plattner Institute, University of Potsdam. My independent research record includes the TOPOLOGIX project, a sequence-based machine learning system that predicts drug-resistance mutations from protein language model embeddings and chemical fingerprints. On the Platinum benchmark of 553 mutations, TOPOLOGIX achieves an AUROC of 0.804 plus or minus 0.025, outperforming the structure-based baseline mCSM-lig at approximately 0.70 while covering 100 percent of mutations compared to roughly 18 percent for structure-limited tools. This matters for Nigeria because most clinically relevant resistance mutations occur in proteins with no experimentally determined structure. A sequence-only method works where structure-based tools fail. OWSD's selection criteria emphasize academic excellence, a sound research plan, and demonstrated commitment to returning home. My academic record is documented. My research plan, detailed in the attached statement, is pre-registered, falsifiable, and calibrated to African resistance data. My commitment to Nigeria is evidenced by my employment history: clinical pharmacist at Ramset Pharmacy, research assistant at the Centre for Drug Discovery, Development and Production, and bioinformatics researcher with the Genomic Health Research Unit on antimicrobial resistance genomics. I am seeking training and resources to return with tools that work in Nigerian conditions, not a fellowship to leave Nigeria permanently. I am 29 years old, a Nigerian citizen, and I have no alternative funding source for a PhD. The OWSD fellowship is my primary path to doctoral training. I ask for your consideration. RESEARCH STATEMENT Project Title: Sequence-Based Prediction of Drug Resistance Mutations for African Antimicrobial Resistance Surveillance The problem. Antimicrobial resistance surveillance in Africa relies on whole-genome sequencing, but the interpretation bottleneck is severe. When a clinician or public health lab obtains a genome sequence, they need to know which mutations confer resistance to which drugs. Current tools such as mCSM-lig require a protein structure, and structures exist for only a small fraction of resistance-relevant proteins. My prior work quantified this gap: on the Platinum benchmark, structure-limited tools cover approximately 18 percent of mutations. The remaining 82 percent are invisible to them. The method. TOPOLOGIX, my current system, uses ESM-2 protein language model delta-embeddings to represent the mutation's effect on the protein sequence, Morgan/ECFP fingerprints to represent the drug, and a Random Forest classifier to predict resistance. On the Platinum benchmark of 553 mutations, it achieves AUROC 0.804 plus or minus 0.025, beating the structure-based mCSM-lig baseline at approximately 0.70. On SKEMPI 2.0, it achieves 0.634. The system covers 100 percent of mutations because it requires only sequence and chemical structure, both of which are always available. The PhD project. The proposed doctoral research extends TOPOLOGIX in three directions, each chosen for direct relevance to African resistance data. First, domain adaptation. TOPOLOGIX was trained and evaluated on Platinum, a benchmark dominated by Western resistance data. African resistance mutations differ in their distribution and context. I will assemble a dataset of resistance mutations from Nigerian and West African genomic surveillance studies, including data from the Genomic Health Research Unit where I previously worked, and evaluate TOPOLOGIX's performance on this distribution. If performance degrades, I will fine-tune the model on African data and document the transfer learning requirements. Second, multi-drug combination prediction. Clinical resistance is rarely a single mutation against a single drug. I will extend the model to predict resistance profiles across drug combinations, using the same sequence and fingerprint representations but with a multi-label output layer. This directly addresses the clinical question in Nigerian tuberculosis and Gram-negative infection management, where combination therapy is standard. Third, uncertainty quantification. A prediction without a confidence interval is not actionable in a clinical setting. I will implement conformal prediction or Bayesian neural network uncertainty estimates so that each prediction carries a calibrated reliability score. This is a methodological contribution that also has immediate practical value for clinicians deciding whether to trust a prediction. Validation plan. The project is pre-registered. The primary endpoint is AUROC on a held-out African resistance mutation dataset, with a pre-specified threshold for clinical utility. Secondary endpoints include coverage, calibration of uncertainty estimates, and comparison against structure-based baselines where structures are available. All code, data, and pre-registration documents will be published openly on GitHub and Zenodo, consistent with my prior practice across five pre-registered studies. Feasibility. I have the computational infrastructure skills: Python, PyMC for Bayesian calibration, RDKit for chemical fingerprints, and experience with HPC via SLURM and Nextflow. I have domain expertise as a licensed pharmacist and prior bioinformatics experience in AMR genomics. The Hasso Plattner Institute provides the computational environment for the first phase of the work. The project requires no wet lab, no animal facilities, and no equipment beyond standard compute, which makes it feasible in both Germany and Nigeria. Alignment with OWSD. This is a natural-sciences and IT project with direct public health relevance to an STLC. It builds on my existing independent research record, including a pre-registered replication that falsified a published claim about topological descriptors for cardiotoxicity prediction, which demonstrates the methodological rigor OWSD seeks. The project trains me in machine learning, protein bioinformatics, and clinical translation, and it produces an open-source tool that Nigerian public health laboratories can deploy without proprietary software. The return commitment is built into the research design. SHORT ESSAY: LEADERSHIP AND IMPACT My leadership is demonstrated through independent research execution rather than formal titles. I have designed, pre-registered, and completed five research studies as a sole or lead author, including the CCT model of reward-memory encoding in addiction, which confirmed all five pre-registered hypotheses with posterior super-additivity of 13 to 22 percentage points across model versions, and the TOPOLOGIX resistance prediction system described in my research statement. Each of these required managing the full research lifecycle: literature review, method design, implementation, statistical analysis, and open publication. My commitment to returning to Nigeria is the design principle of my career. I have worked as a clinical pharmacist at Ramset Pharmacy, as a research assistant at the Centre for Drug Discovery, Development and Production where I performed NMDA and insulin docking studies, and as a bioinformatics researcher with the Genomic Health Research Unit on AMR surveillance pipelines. These positions gave me direct experience with the infrastructure gaps that OWSD fellows are expected to address. I know what it means to run a surveillance pipeline on unreliable power and intermittent internet. I know which tools fail in those conditions and which do not. The impact I intend to have is concrete. Within five years of completing the PhD, I will establish a computational resistance-prediction service at a Nigerian university or research institute, trained on local data and deployed for local clinicians. The TOPOLOGIX system, extended and validated through this fellowship, will be the core of that service. I will also train graduate students in sequence-based machine learning methods, transferring the skills I gain at the Hasso Plattner Institute to the next generation of Nigerian computational biologists. This is the capacity-building outcome OWSD exists to produce, and it is the outcome I am positioned to deliver. SHORT ESSAY: FINANCIAL NEED I am applying for the OWSD PhD Fellowship because I have no alternative funding source for doctoral training. My current enrolment in the M.Sc. Digital Health programme at the Hasso Plattner Institute is self-funded through my employment as National Product Manager at Synthcare and prior clinical pharmacy work. These earnings cover tuition and living costs for the master's degree but do not extend to a multi-year PhD programme. I have not received research grants or fellowships previously. My research to date has been conducted independently, using open-source tools and publicly available datasets, which is why my publication record consists of preprints on OSF and Zenodo rather than paywalled journals. The OWSD fellowship would cover the direct costs of doctoral research: tuition, living expenses, and computational resources. It would also provide the institutional legitimacy that facilitates collaboration with Nigerian public health agencies and academic partners. I am 29 years old, within the age limit, and a citizen of Nigeria, which is on the OWSD list of Science and Technology Lagging Countries. I meet the eligibility criteria and I lack the financial resources to pursue this PhD without fellowship support. CHECKLIST - [ ] Verify current OWSD PhD Fellowship eligibility criteria on the official OWSD website, including age limit, degree requirements, and STLC citizenship list - [ ] Confirm that Nigeria is on the current OWSD STLC list and that the Hasso Plattner Institute or a partner university qualifies as a recognized institution in a developing country - [ ] Obtain official admission or enrolment letter from the PhD programme at Hasso Plattner Institute or an alternative host university - [ ] Request three letters of recommendation from academic referees, preferably including one from a supervisor at the Genomic Health Research Unit and one from a professor at the University of Ibadan - [ ] Prepare certified copies of B.Pharm degree certificate and transcript from the University of Ibadan - [ ] Prepare current enrolment certificate from the Hasso Plattner Institute M.Sc. Digital Health programme - [ ] Prepare PCN pharmacy license certificate - [ ] Update CV to include all research preprints with DOIs from OSF and Zenodo - [ ] Prepare pre-registration documents for the proposed PhD project and upload to OSF - [ ] Prepare a data management and open science plan for the TOPOLOGIX extension project - [ ] Draft a budget table for the fellowship application, including tuition, living costs, compute, and conference travel - [ ] Verify the application deadline and submission portal on the OWSD website, as the MaPage listing does not specify a deadline - [ ] Confirm whether the application requires a host supervisor letter of support from the PhD institution - [ ] Prepare a timeline for the PhD project, including milestones for dataset assembly, model extension, and validation EDITOR NOTES - Eligibility risk: The OWSD PhD Fellowship requires the applicant to be admitted to or enrolled in a PhD programme at a recognized university in a developing country outside her home country. Eniola is currently enrolled in an M.Sc. at Hasso Plattner Institute in Germany, which is not a developing country. She must confirm whether the PhD host institution can be in Germany or whether she needs to identify a host university in another developing country. This is the single largest eligibility risk and must be resolved before submission. - Verification needed: The MaPage listing provides no deadline, amount, or provider details beyond the OWSD name. The applicant must check the official OWSD website for the current call, as criteria and deadlines change annually. The age limit of under 40 is noted in the strategy notes but must be confirmed against the current call text. - Gap to fill: The profile does not specify which PhD programme or host university Eniola will attend. The research statement assumes Hasso Plattner Institute as the host, but the OWSD requirement for a developing-country host institution may force a different choice. The applicant must insert the actual host university and supervisor name before submission. - Personal detail needed: The leadership essay references a commitment to establishing a resistance-prediction service at a Nigerian institution but does not name the institution. The applicant should insert a specific target institution, ideally one where she has existing contacts, such as the University of Ibadan or the Genomic Health Research Unit. - Consistency check: The research statement correctly leads with TOPOLOGIX as the core project, consistent with the recommended framing angle. The cardiotoxicity topology study is mentioned only as evidence of methodological rigor, which is appropriate. The CCT model, neurocascade, ergofluids, and psyche-twin are not mentioned in the application materials, which is correct for this programme's focus on natural sciences and IT with public health relevance.