MOTIVATION LETTER The Deutschland Stipendium rewards academic excellence and social commitment among enrolled students at German universities. My enrollment in the M.Sc. Digital Health program at the Hasso Plattner Institute, University of Potsdam, beginning Winter Semester 2026/27, places me within the programme's eligible cohort. The stipend's dual mandate, academic merit and social engagement, matches the two pillars of my work: peer-reviewed computational research and open, reproducible science that serves African research communities. My research record includes three sole-authored preprints under review at peer-reviewed journals (IART, PNPBP, NBR) and one co-authored paper under review at Alcohol (Elsevier). The CCT model, a tripartite pharmacological framework for reward-memory encoding prevention in addiction, confirmed all five pre-registered hypotheses with posterior super-additivity of 13-22 percentage points across model versions. The TOPOLOGIX project, my current focus, applies ESM-2 protein-language-model delta-embeddings with Morgan fingerprints to predict drug-resistance mutations from sequence alone, achieving AUROC 0.804 plus or minus 0.025 on the Platinum benchmark (553 mutations) and covering 100 percent of mutations versus roughly 18 percent for structure-limited tools. This work addresses antimicrobial resistance, a public health crisis that disproportionately affects Nigeria and other LMICs. Social engagement is embedded in how I work. I publish pre-registered protocols on OSF, deposit data on Zenodo, and maintain an open GitHub repository (github.com/AmunRaPtah) with reproducible pipelines. My employment history includes a Bioinformatics Researcher role with GHRU-GSAR, where I contributed to AMR genomics surveillance pipelines that inform public health responses in Nigeria. As a licensed pharmacist (PCN) with a B.Pharm from the University of Ibadan, I have direct clinical experience with the consequences of drug resistance. The stipend would allow me to reduce paid pharmacy hours and dedicate that time to completing TOPOLOGIX and publishing its results in open-access venues. The €300 monthly stipend is modest, but it is precisely calibrated to my situation. It covers a meaningful portion of living costs in Potsdam, enabling full focus on coursework and research rather than part-time work. The Deutschland Stipendium's selection criteria, academic achievement and social commitment, are criteria I meet with verifiable evidence. I request consideration on the strength of that evidence. RESEARCH STATEMENT The TOPOLOGIX project addresses a specific, measurable gap in drug-resistance prediction: structure-based tools fail for the majority of clinically relevant mutations because protein structures are unavailable. mCSM-lig, a widely used structure-based predictor, covers roughly 18 percent of mutations in the Platinum benchmark. TOPOLOGIX covers 100 percent by using sequence alone. The method combines ESM-2 protein-language-model delta-embeddings, which capture the effect of a mutation on the learned representation of the protein sequence, with Morgan/ECFP drug fingerprints and a Random Forest classifier. The model predicts whether a given mutation confers resistance to a given drug. On the Platinum benchmark (553 mutations), TOPOLOGIX achieves AUROC 0.804 with a standard deviation of 0.025. On SKEMPI 2.0, it achieves 0.634. These results beat structure-based baselines while requiring no structural data. This project began with a falsified hypothesis. I tested whether bipartite persistent homology, an opposition-distance metric computed with Ripser and GUDHI, could predict 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 then applied the same topological constructs to drug-resistance prediction and found they carry almost no signal (AUROC 0.425 and 0.485 on the Platinum benchmark). These negative results, reported directly rather than reframed, ruled out interface geometry as the driver and motivated the sequence-representation approach that became TOPOLOGIX. The current stage is validation and extension. The model's performance on Platinum is strong, but SKEMPI 2.0 performance (0.634) indicates room for improvement. Planned work includes incorporating attention-based pooling over ESM-2 embeddings, testing transfer across drug classes, and benchmarking against additional resistance databases. All code, data splits, and pre-registered protocols are available on GitHub and OSF. The relevance to the Deutschland Stipendium is direct. The stipend supports enrolled students with demonstrated academic excellence. TOPOLOGIX is my primary academic output during the M.Sc. program. The stipend would fund the time needed to complete the extension work and prepare manuscripts for submission. The societal stakes are concrete: antimicrobial resistance is projected to cause 10 million deaths annually by 2050, with the highest burden in Africa. A sequence-only predictor that works where structure-based tools fail is a practical contribution to that fight. EDITOR NOTES - Research line selected: TOPOLOGIX. Rationale: it is the current project, it demonstrates both academic excellence (AUROC 0.804 on Platinum, beating structure-based baselines) and social engagement (antimicrobial resistance, a crisis concentrated in Nigeria and other LMICs), and it is the most direct match to the stipend's merit-plus-social-commitment criteria. The CCT model is strong but is addiction neuroscience, which maps less cleanly to the stipend's social-engagement pillar. The hERG and resistance topology studies are included as falsified results that motivated TOPOLOGIX, presented honestly as superseded work, not current claims. - Eligibility risk: the Deutschland Stipendium is awarded to enrolled students. The applicant is enrolled for Winter Semester 2026/27 at HPI/Potsdam. Confirm enrollment is active at the time of application and that the stipend's eligibility window covers the full M.Sc. period. The application deadline is listed as "see programme website"; verify the current cycle's deadline and whether enrollment must be confirmed before or after application. - Facts to verify: the AUROC figures for TOPOLOGIX (0.804 plus or minus 0.025 on Platinum, 0.634 on SKEMPI 2.0), the coverage figures (100 percent versus roughly 18 percent for mCSM-lig), and the mCSM-lig baseline (approximately 0.70) must be checked against the current preprint versions before submission. The claim that all five CCT hypotheses were confirmed should be verified against the latest preprint. The Alcohol (Elsevier) co-authored paper's status should be confirmed as under review. - Gaps to fill: the motivation letter references "reducing paid pharmacy hours" but does not specify current employment status at the time of application. The applicant's profile lists a National Product Manager role at Synthcare from March 2026; clarify whether this role continues during the M.Sc. program and how the stipend would interact with it. The social-engagement section should include at least one concrete, verifiable example of mentorship or community work in Nigeria, as the current draft implies it but does not name a specific activity, organization, or outcome. - Formatting note: the Deutschland Stipendium application is typically submitted through the university's online portal, not through DAAD directly. Confirm the submission route for HPI/Potsdam specifically, as the portal link in the programme description is a general DAAD database entry. The required materials may include a CV, transcript, and proof of enrollment in addition to the motivation letter; confirm the exact list before submission.