MOTIVATION LETTER The TOPOLOGIX project began with a falsified hypothesis. My pre-registered, powered replication on hERG cardiotoxicity proved that bipartite persistent homology of protein-ligand interfaces does not beat a plain descriptor baseline (AUROC 0.8426 versus 0.8782). A second study on the Platinum benchmark confirmed interface geometry carries almost no signal for drug-resistance prediction (AUROC 0.425 and 0.485). Those negative results cleared the ground for what works: sequence alone. TOPOLOGIX now predicts drug-resistance mutations from ESM-2 protein-language-model delta-embeddings plus Morgan fingerprints, reaching AUROC 0.804 plus or minus 0.025 on Platinum (553 mutations) and 0.634 on SKEMPI 2.0, while covering 100 percent of mutations versus roughly 18 percent for structure-limited tools like mCSM-lig at 0.70. This is the research I want to deepen at Universite Paris-Saclay. The International Master's Scholarships programme selects for academic excellence, research potential, and clarity of academic project. My B.Pharm from the University of Ibadan (CGPA 5.1/7.0, German equivalent 1.9) and my current enrollment in the M.Sc. Digital Health at Hasso Plattner Institute establish the academic record. My three sole-authored preprints under review at peer-reviewed journals (IART, PNPBP, NBR), a co-authored paper under review at Alcohol (Elsevier), and the TOPOLOGIX results demonstrate independent, publishable research. I am 29 years old, a Nigerian national who has never resided in France, and I am applying as a newly arriving international student. I meet every stated eligibility condition. Paris-Saclay is the correct environment for the next stage of this work. The university's concentration of computational biology, AI/ML for health, and protein engineering groups matches TOPOLOGIX's method stack: protein language models, Bayesian calibration, and dynamical-systems thinking. My Bayesian MCMC calibration experience (PyMC DEMetropolisZ, 14 free parameters, literature-elicited priors from a 1,847-record screen) and my circuit-level pharmacology simulation engine neurocascade (62 of 62 tests passing, three receptor systems calibrated) are directly relevant to the computational neuroscience and AI-for-health programmes at Paris-Saclay. I am looking for the research training and institutional context that will position me for a PhD in computational neuroscience or protein engineering, and Paris-Saclay is where that training is strongest. The scholarship is merit-based, not need-based, and that is precisely the basis on which I apply. My record shows a researcher who designs pre-registered studies, reports negative results honestly, and pivots to methods that work. TOPOLOGIX is the current, validated line of that work. Paris-Saclay is where I want to take it next. RESEARCH STATEMENT My research program asks one question across scales: how do molecular interactions produce system-level outcomes, and which representations of those interactions actually predict the outcomes we care about? The answer, from my work so far, is that the most geometrically intuitive representations are often the least predictive, and that sequence-based and dynamical representations carry more signal than structure-based ones. The CCT (Conjunctive Consolidation Threshold) model addresses addiction at the circuit level. It is a tripartite pharmacological framework for reward-memory encoding prevention, coupling dopaminergic reward prediction error, NMDAR-dependent long-term potentiation, and affective contrast in a three-axis ODE model solved with RK45. I calibrated all 14 free parameters with Bayesian MCMC against literature-elicited priors from a screen of 1,847 records. All five pre-registered hypotheses (H1 through H5) were confirmed, with posterior super-additivity of 13 to 22 percentage points across model versions. Three sole-authored preprints are in review at peer-reviewed journals. The cardiotoxicity and resistance studies tested whether bipartite persistent homology of protein-ligand interfaces predicts biological outcomes. The results were negative and decisive: AUROC 0.8426 versus 0.8782 for the descriptor baseline on hERG cardiotoxicity, and 0.425 and 0.485 on the Platinum benchmark for drug resistance. These studies settled comparisons the published literature had never actually run, and they redirected my work toward sequence representations. TOPOLOGIX is that redirection. It uses ESM-2 protein-language-model delta-embeddings combined with Morgan/ECFP drug fingerprints and a Random Forest classifier to predict drug-resistance mutations from sequence alone. On the Platinum benchmark (553 mutations), TOPOLOGIX achieves AUROC 0.804 plus or minus 0.025; on SKEMPI 2.0, 0.634. It beats structure-based baselines (mCSM-lig at approximately 0.70) while covering 100 percent of mutations versus roughly 18 percent for structure-limited tools. This is the project I propose to develop at Paris-Saclay. The methodological thread through all of this work is honest validation. Every study is pre-registered. Every negative result is reported directly rather than reframed. The ergofluids project, which extends Koopman-operator and Dynamic Mode Decomposition methods with a Mori-Zwanzig memory kernel for macromolecular transport in tumor tissue, passed its synthetic-data gates but did not meet its primary pre-registered criterion on the first real-data gate. That result is reported as it stands. Methods-validation research does not get to cherry-pick its outcomes. At Paris-Saclay, I want to extend TOPOLOGIX in two directions. First, move from Random Forest to deeper sequence models and test whether the delta-embedding approach generalizes across protein families beyond the Platinum benchmark. Second, connect TOPOLOGIX's sequence-level predictions to the circuit-level dynamics of the CCT model, building a pipeline from mutation to molecular mechanism to systems-level consequence. Paris-Saclay's computational biology and AI/ML for health groups have the expertise and infrastructure for both directions. My Bayesian modeling skills, my HPC experience (Nextflow, SLURM), and my track record of independent, pre-registered research make me ready to contribute immediately. EDITOR NOTES - Research line selected: TOPOLOGIX, because it is the current, validated project with the strongest direct match to Paris-Saclay's computational biology and AI strengths. The CCT model is mentioned as supporting evidence of Bayesian and dynamical-systems skill, not as the primary pitch. The falsified topology results are presented as motivation for TOPOLOGIX, which is their honest role. - Eligibility risk: the profile states Eniola is enrolled in M.Sc. Digital Health at HPI/Potsdam for Winter Semester 2026/27. The Paris-Saclay scholarship requires newly arriving students not already enrolled in a Paris-Saclay programme. Verify whether simultaneous enrollment in another master's programme is permitted or whether the HPI enrollment must be deferred or withdrawn. This must be confirmed before submission. - Age verification: the profile states age 29. The scholarship requires under 30 at time of application. Confirm the exact application deadline and calculate age against it. If the birthday falls before the deadline, the application is ineligible. - The motivation letter and research statement assume the applicant is applying to a specific Paris-Saclay master's programme (e.g., M.Sc. in Computational Neuroscience or AI/ML for Health). The actual programme name and its application form must be inserted, and the letter may need to name the specific programme explicitly. - The profile lists employment through March 2026 and enrollment starting Winter Semester 2026/27. The Paris-Saclay application timeline must be checked against these dates to ensure the narrative of "newly arriving" is factually consistent. - All numerical claims (AUROC values, coverage percentages, test counts, CGPA) are drawn directly from the profile and should be verified against the actual preprints before submission. CHECKLIST - [ ] Confirm eligibility: age under 30 at application deadline, not resided in France more than 1 year in past 5 years, newly arriving student - [ ] Verify HPI/Potsdam enrollment status does not conflict with Paris-Saclay "newly arriving" requirement - [ ] Identify and select the specific Paris-Saclay M.Sc. programme (e.g., Computational Neuroscience, AI/ML for Health) - [ ] Complete the university admissions application for the selected M.Sc. programme - [ ] Complete the International Master's Scholarships application form - [ ] Upload motivation letter (300-500 words, final version) - [ ] Upload research statement (400-600 words, final version) - [ ] Upload academic transcripts (B.Pharm, University of Ibadan; current HPI enrollment proof) - [ ] Upload CV listing publications, preprints, and employment history - [ ] Upload ORCID record (0009-0001-9272-6735) and GitHub profile (github.com/AmunRaPtah) - [ ] Upload preprints: three sole-authored CCT papers, co-authored Alcohol paper, TOPOLOGIX results - [ ] Obtain and upload letters of recommendation (Berridge, Gershman, Daw, or Mattar) - [ ] Verify all numerical claims against preprints before final submission - [ ] Submit before the programme deadline (check website for exact date)