MOTIVATION LETTER The Gitcoin DeSci round exists to fund research that refuses to hide behind paywalls and closed review processes. My work as an independent computational researcher in Nigeria has been built on that same premise since 2024. I have published three sole-authored preprints on OSF and Zenodo, each currently in review at a peer-reviewed journal, covering the Conjunctive Consolidation Threshold model for addiction pharmacology. I have reported negative results directly when my pre-registered hypotheses failed, as in my cardiotoxicity topology study where bipartite persistent homology scored an AUROC of 0.8426 against a plain descriptor baseline of 0.8782. That comparison had never been run in the published literature, and I ran it, reported it, and moved on. That is decentralized science in practice: open methods, open data, open failures. My current project, TOPOLOGIX, predicts drug-resistance mutations from protein language model embeddings and drug fingerprints. It achieves an AUROC of 0.804 plus or minus 0.025 on the Platinum benchmark of 553 mutations, and 0.634 on SKEMPI 2.0. It beats structure-based baselines like mCSM-lig at roughly 0.70 while covering 100 percent of mutations, compared to about 18 percent for structure-limited tools. The code is on GitHub. The benchmarks are public. The pre-registration is on OSF. This is exactly the kind of open, reproducible computational tool that DeSci infrastructure should host and fund. What I need from Gitcoin is not just funding, though the grant matters. I need the Web3 layer that my current workflow lacks. My models and preprints live on centralized servers. My validation pipelines are reproducible but not yet decentralized. I propose to build an IPFS-hosted model registry for my CCT and TOPOLOGIX outputs, with a DAO-governed validation mechanism where community members can stake on replication attempts. I will use Gitcoin's community governance model as the template for how that registry makes decisions. The budget, roughly 8,000 USD, covers IPFS storage, a smart contract audit, and one month of dedicated development time. The remaining funds support my ongoing M.Sc. in Digital Health at the Hasso Plattner Institute, where I will integrate these decentralized tools into my coursework. My perspective as a Nigerian researcher matters here. DeSci cannot claim global reach if African researchers are only consumers of infrastructure built elsewhere. I am building the infrastructure. I have already deployed four independent DuckDB-based ingest-to-analyze pipelines across life sciences, tech, and social science domains, and I self-host local LLM serving with llama.cpp on my own Linux VPS. I know how to run production systems. I know how to report results honestly. I am asking Gitcoin to fund the next layer. RESEARCH STATEMENT The Conjunctive Consolidation Threshold model is a tripartite pharmacological framework for preventing reward-memory encoding in addiction. It couples three axes: dopaminergic reward prediction error, NMDAR-dependent long-term potentiation, and affective contrast. The model is implemented as a system of ordinary differential equations solved with RK45, calibrated with Bayesian MCMC using PyMC's DEMetropolisZ sampler across 14 free parameters. Priors were elicited from a systematic screen of 1,847 records in the literature. All five pre-registered hypotheses, H1 through H5, were confirmed. Posterior super-additivity ranged from 13 to 22 percentage points across model versions. The model is described in three sole-authored preprints, each under review at a different journal: IART, PNPBP, and NBR. A co-authored paper is under review at Alcohol, Elsevier. The CCT model is open source. The code is on GitHub. The preprints are on OSF and Zenodo. The priors, the calibration runs, and the posterior distributions are all published alongside the text. This is the standard I hold myself to, and it is the standard DeSci exists to enforce. My second active line is TOPOLOGIX, which predicts drug-resistance mutations from sequence alone. The method uses ESM-2 protein language model delta-embeddings combined with Morgan and ECFP drug fingerprints, classified by a Random Forest. On the Platinum benchmark of 553 mutations, it achieves an AUROC of 0.804 plus or minus 0.025. On SKEMPI 2.0, it achieves 0.634. Structure-based baselines such as mCSM-lig score around 0.70 but cover only about 18 percent of mutations because they require resolved structures. TOPOLOGIX covers 100 percent. This matters for antimicrobial resistance surveillance in low-resource settings, where structural data is rarely available and sequence data is increasingly cheap to generate. I have also run two studies that produced negative results, and I report them here because DeSci values honest negative findings. My cardiotoxicity topology study tested whether bipartite persistent homology predicts hERG cardiotoxicity from protein-ligand interface geometry. It does not beat a plain descriptor baseline. My interface-topology-for-resistance study found that the same topological constructs carry almost no signal for drug-resistance prediction, with AUROC scores of 0.425 and 0.485 on the Platinum benchmark. These results ruled out a hypothesis class and redirected my work toward sequence representations, which produced TOPOLOGIX. My broader toolkit includes neurocascade, a receptor-to-behavior brain-circuit simulation engine that couples pharmacokinetics to receptor binding to Wilson-Cowan circuit dynamics to behavioral readouts, with 62 of 62 tests passing. I also maintain ergofluids, a Koopman-operator and Dynamic Mode Decomposition project with a Mori-Zwanzig memory kernel for modeling drug transport through tumor tissue, where I reported a failed real-data gate directly rather than reframing it. For the Gitcoin DeSci round, I propose to make these models the foundation of a decentralized repository. The repository will host model outputs, calibration data, and validation pipelines on IPFS. A DAO-governed mechanism will allow community members to propose replication studies and vote on validation criteria. This turns my existing open-science workflow into a Web3-native infrastructure that other researchers, especially in Africa, can reuse without asking permission. BUDGET BREAKDOWN Total requested: 8,000 USD - IPFS storage and pinning services for model outputs, calibration data, and preprints: 1,200 USD - Smart contract audit for the DAO-governed validation registry: 2,000 USD - Development time for the registry interface and IPFS integration: 3,000 USD - Travel and conference fees to present the decentralized repository at a DeSci or neuroscience meeting: 800 USD - Open-access publication fees for one CCT-related manuscript: 1,000 USD TEAM BIO Eniola Ayodele Olutogun is an independent computational researcher based in Nigeria, holding a B.Pharm from the University of Ibadan with a German equivalent grade of 1.9. He is a PCN-licensed pharmacist and is enrolled in the M.Sc. Digital Health program at the Hasso Plattner Institute and University of Potsdam, starting Winter Semester 2026/27. His research spans addiction neuroscience, protein machine learning, and dynamical-systems methods. He has received endorsements from Kent Berridge at Michigan, Samuel Gershman at Harvard, Nathaniel Daw at Princeton, and Marcelo Mattar at NYU. His ORCID is 0009-0001-9272-6735 and his GitHub is github.com/AmunRaPtah. COMMUNITY ENGAGEMENT PLAN The decentralized repository will be open for community contribution from day one. I will publish a contribution guide alongside the IPFS-hosted registry, specifying how researchers can submit replication attempts, new benchmark datasets, or model extensions. I will run a monthly community call to review open validation proposals and discuss results. I will also integrate the repository with Gitcoin's existing grant infrastructure so that community members can propose micro-grants for specific replication studies. My existing open-science record, including pre-registrations on OSF and public negative-result reporting, demonstrates that I will follow through on these commitments. CHECKLIST - [ ] Confirm Gitcoin round eligibility at https://checker.gitcoin.co/public/round/ke48A6DoM/eligibility - [ ] Verify that all GitHub repositories are public and linked to the application - [ ] Confirm OSF and Zenodo preprint links are active and accessible - [ ] Verify ORCID record is up to date with all preprints and affiliations - [ ] Draft and upload the motivation letter as a PDF - [ ] Draft and upload the research statement as a PDF - [ ] Draft and upload the budget breakdown as a PDF - [ ] Draft and upload the team bio as a PDF - [ ] Draft and upload the community engagement plan as a PDF - [ ] Confirm wallet address for Gitcoin grant distribution - [ ] Verify enrollment status at Hasso Plattner Institute for the M.Sc. Digital Health program - [ ] Confirm all five pre-registered hypotheses H1-H5 are documented in the CCT preprints - [ ] Confirm AUROC values for TOPOLOGIX are stated correctly in the application - [ ] Submit the application before the rolling deadline EDITOR NOTES - Eligibility risk: The Gitcoin DeSci round focuses on Web3 infrastructure and decentralized tools. The applicant's core research is computational neuroscience and protein ML. The proposal must clearly frame the IPFS-hosted registry and DAO governance as the Web3 layer, not the research itself. If the reviewers expect a fully decentralized application, the registry proposal may need to be more technically detailed. - Verification needed: The enrollment status at Hasso Plattner Institute for Winter Semester 2026/27 should be confirmed with an official letter or admission notice. The application should not assume the reader knows the German academic calendar. - Verification needed: The budget breakdown totals 8,000 USD, but the round amount is unspecified. The applicant should confirm whether this falls within the round's typical grant range. If the round funds smaller amounts, the budget should be scaled down. - Gap: The community engagement plan mentions monthly community calls but does not specify a platform or time zone. The applicant should insert a concrete plan, such as a Discord server or a monthly Zoom link, and account for the Nigeria time zone. - Gap: The team bio lists endorsements from four senior researchers. The applicant should confirm whether any of these researchers are willing to provide a letter of support for the Gitcoin application. If not, the endorsements should be framed as advisory only.