Draft — NLnet Foundation Grant

MOTIVATION LETTER

NLnet Foundation funds internet infrastructure that serves the public good. The open-source platforms I have built, IMPRINT, TOPOLOGIX, and GATE, are computational tools designed to make addiction and drug-safety research reproducible, transparent, and accessible to any researcher worldwide, regardless of institutional resources. These tools are licensed under Apache 2.0, hosted on GitHub, and already used in my independent research on the Conjunctive Consolidation Threshold (CCT) model of reward-memory encoding. I am applying to the OISI track to harden TOPOLOGIX into a general-purpose topological data analysis library for drug safety screening, and to extend GATE into a reusable BCI neural-stimulation safety evaluation framework.

TOPOLOGIX currently uses persistent homology and bipartite simplicial complexes to predict hERG cardiotoxicity from protein-drug interaction data. The MVP is functional and validated against known cardiotoxic compounds. With NLnet support, I will add a standardized API, thorough documentation, unit tests, and a plugin architecture that allows researchers to plug in their own molecular descriptors or protein structures. The deliverable is a pip-installable Python package with a DOI on Zenodo, a peer-reviewed software paper, and a tutorial notebook series targeting pharmacology and neuroscience labs in low-resource settings.

GATE evaluates safety parameters for BCI neural-stimulation protocols. It currently runs as a command-line tool with NEURON and Brian2 backends. NLnet funding would allow me to containerize the tool with Docker, add a web-based configuration interface, and publish a benchmark dataset of simulated stimulation outcomes. The public-benefit angle is direct: GATE enables any lab to pre-screen stimulation parameters before animal or human trials, reducing harm and lowering the barrier to entry for BCI research in Africa.

The CCT model provides the scientific motivation for these tools, but the grant request is for software infrastructure. I am asking NLnet to fund the open-source computational plumbing that makes that research reproducible and scalable. My track record includes three sole-authored preprints, a provisional patent on the CCT core architecture, endorsements from Kent Berridge, Samuel Gershman, Nathaniel Daw, and Marcelo Mattar, and functional code repositories for all three platforms. I am an independent researcher based in Lagos, Nigeria, with no institutional affiliation, and I am building these tools because the existing commercial and academic alternatives are closed, expensive, or unavailable in my region.

The requested amount is EUR 35,000 over 12 months. Deliverables include a stable release of TOPOLOGIX v1.0 with documentation and test suite, a containerized GATE v2.0 with web interface, a published software paper, and a workshop for African computational neuroscience researchers on using these tools. All code will remain Apache 2.0.

RESEARCH STATEMENT

My independent research program centers on the Conjunctive Consolidation Threshold (CCT) model, a tripartite pharmacological framework for preventing reward-memory encoding in addiction. The model specifies that three concurrent pharmacological actions, dopamine D1 receptor antagonism, NMDA receptor partial agonism, and mu-opioid receptor antagonism, must exceed a conjunctive threshold to prevent the consolidation of reward-associated memories. I have validated this model through ODE/RK45 simulations and Bayesian MCMC population dynamics, achieving an encoding probability reduction from 0.855 to 0.122 (85.8% reduction) and demonstrating super-additivity of 12.8 percentage points beyond the sum of individual effects. All five pre-registered hypotheses (H1-H5) were confirmed. The formal mathematical specification is published on OSF (10.17605/OSF.IO/EMY4U), and the Bayesian clinical trial architecture is on Zenodo (10.5281/zenodo.20492472). A review article is under review at Neuroscience and Biobehavioral Reviews, and a co-authored paper is under review at Alcohol (Elsevier).

The CCT model is not directly fundable by NLnet. However, the computational tools I built to develop and test the model are precisely the kind of open-source infrastructure NLnet supports. IMPRINT screens compounds for addiction liability using the CCT framework. TOPOLOGIX uses topological data analysis to predict drug-protein interactions and cardiotoxicity. GATE evaluates BCI neural-stimulation safety. All three are Apache 2.0, modular, and designed for reproducibility. The NLnet grant would allow me to generalize TOPOLOGIX beyond hERG cardiotoxicity into a full TDA library for drug safety screening, and to harden GATE into a reusable framework with containerized deployment and a web interface.

The broader impact is that any addiction researcher, pharmacologist, or BCI engineer, especially those in Africa, South Asia, or Latin America, can use these tools without paying for licenses or needing high-performance computing clusters. I am building the infrastructure I needed but did not have when I started this work as an independent researcher in Lagos.

BUDGET NARRATIVE

Total request: EUR 35,000 over 12 months.

Personnel: EUR 20,000 for 12 months of full-time development work by the applicant (Eniola Ayodele Olutogun). This is a subsistence-level stipend for an independent researcher in Lagos, Nigeria, covering rent, food, internet, and electricity. No other personnel are requested.

Computing: EUR 5,000 for cloud compute credits (AWS or Hetzner) for running TDA pipelines, Bayesian MCMC sampling, and container builds. This covers approximately 10,000 CPU-hours and 500 GPU-hours.

Software dependencies and licensing: EUR 2,000 for DOI registration, software deposit fees (Zenodo, Software Heritage), and any required licenses for third-party libraries that are not fully open-source.

Workshop and dissemination: EUR 5,000 for hosting a two-day virtual workshop for African computational neuroscience researchers, including speaker honoraria (EUR 500 each for three invited speakers), platform costs, and recording/transcription. Remaining EUR 3,000 for travel to one international conference (e.g., Computational and Systems Neuroscience, COSYNE) to present the tools and recruit beta testers.

Contingency: EUR 3,000 (approximately 8.6%) for unforeseen costs, exchange rate fluctuations, or extended compute time.

All funds will be managed through a personal account in Nigeria. No institutional overhead is charged.

TIMELINE AND MILESTONES

Month 1-2: Audit existing codebases for TOPOLOGIX and GATE. Write unit tests for all core functions. Publish current versions on Zenodo with DOIs. Begin documentation in Sphinx.

Month 3-4: Implement standardized API for TOPOLOGIX. Add support for user-supplied molecular descriptors and protein structures. Write tutorial notebooks for three use cases: hERG cardiotoxicity screening, dopamine receptor binding prediction, and opioid receptor selectivity analysis.

Month 5-6: Containerize GATE with Docker. Add web-based configuration interface using Flask or FastAPI. Publish benchmark dataset of simulated BCI stimulation outcomes (100 parameter combinations, 10,000 simulated trials).

Month 7-8: Beta release of TOPOLOGIX v0.9 and GATE v2.0. Recruit 5-10 beta testers from the computational neuroscience community (contacts via Berridge, Gershman, Daw, Mattar labs). Collect feedback and fix bugs.

Month 9-10: Final release of TOPOLOGIX v1.0 and GATE v2.0. Write and submit software paper to the Journal of Open Source Software (JOSS) or similar. Prepare workshop materials.

Month 11: Host two-day virtual workshop for African computational neuroscience researchers. Record all sessions, publish on YouTube under CC-BY license.

Month 12: Final report to NLnet. Deposit all code, documentation, and workshop materials on Zenodo with persistent identifiers. Publish blog post summarizing project outcomes.

CHECKLIST

- [ ] Motivation letter (500 words, written above)
- [ ] Research statement (400-600 words, written above)
- [ ] Budget narrative (written above)
- [ ] Timeline and milestones (written above)
- [ ] CV or resume (applicant must prepare from profile data)
- [ ] Links to GitHub repositories (github.com/AmunRaPtah/IMPRINT, github.com/AmunRaPtah/TOPOLOGIX, github.com/AmunRaPtah/GATE)
- [ ] Links to preprints (OSF 10.17605/OSF.IO/KG7B5, OSF 10.17605/OSF.IO/EMY4U, Zenodo 10.5281/zenodo.20492472)
- [ ] ORCID profile (0009-0001-9272-6735)
- [ ] Proof of open-source licensing (Apache 2.0 for all three platforms)
- [ ] Two letters of recommendation (suggested: Kent Berridge, Samuel Gershman)
- [ ] Provisional patent documentation (CCT core architecture, Q3 2026)
- [ ] Identification document (passport or national ID)
- [ ] Bank account details for fund transfer (Nigerian bank account)

EDITOR NOTES

- Eligibility risk: NLnet typically funds organizations or individuals with a bank account in the EU/EEA. The applicant is based in Nigeria. Confirm whether NLnet accepts applications from Nigerian individuals, or whether a fiscal sponsor is needed. If not, consider applying through a European collaborator or open-source foundation (e.g., Software Freedom Conservancy).
- Verification needed: Confirm that the three platforms (IMPRINT, TOPOLOGIX, GATE) are indeed licensed under Apache 2.0 and that the GitHub repositories are public and contain functional code. The profile mentions them but does not include repository URLs. The applicant must provide these.
- Gap: The profile does not specify the exact amount requested. EUR 35,000 is an estimate based on typical NLnet grant size and the budget narrative. The applicant should confirm this figure aligns with their actual needs and NLnet's typical range (EUR 5,000-50,000).
- Gap: The profile does not mention any prior experience managing grants or budgets. The applicant should be prepared to explain how they will handle financial reporting and accountability as an independent researcher.
- Gap: The workshop for African computational neuroscience researchers is mentioned in the timeline but not detailed in the profile. The applicant should confirm they have contacts or a mailing list to recruit participants, or plan to partner with an existing organization (e.g., African Institute for Mathematical Sciences, or the Society for Neuroscientists of Africa).

MOTIVATION LETTER

NLnet Foundation funds internet infrastructure that serves the public good. The open-source platforms I have built, IMPRINT, TOPOLOGIX, and GATE, are computational tools designed to make addiction and drug-safety research reproducible, transparent, and accessible to any researcher worldwide, regardless of institutional resources. These tools are licensed under Apache 2.0, hosted on GitHub, and already used in my independent research on the Conjunctive Consolidation Threshold (CCT) model of reward-memory encoding. I am applying to the OISI track to harden TOPOLOGIX into a general-purpose topological data analysis library for drug safety screening, and to extend GATE into a reusable BCI neural-stimulation safety evaluation framework.

TOPOLOGIX currently uses persistent homology and bipartite simplicial complexes to predict hERG cardiotoxicity from protein-drug interaction data. The MVP is functional and validated against known cardiotoxic compounds. With NLnet support, I will add a standardized API, thorough documentation, unit tests, and a plugin architecture that allows researchers to plug in their own molecular descriptors or protein structures. The deliverable is a pip-installable Python package with a DOI on Zenodo, a peer-reviewed software paper, and a tutorial notebook series targeting pharmacology and neuroscience labs in low-resource settings.

GATE evaluates safety parameters for BCI neural-stimulation protocols. It currently runs as a command-line tool with NEURON and Brian2 backends. NLnet funding would allow me to containerize the tool with Docker, add a web-based configuration interface, and publish a benchmark dataset of simulated stimulation outcomes. The public-benefit angle is direct: GATE enables any lab to pre-screen stimulation parameters before animal or human trials, reducing harm and lowering the barrier to entry for BCI research in Africa.

The CCT model provides the scientific motivation for these tools, but the grant request is for software infrastructure. I am asking NLnet to fund the open-source computational plumbing that makes that research reproducible and scalable. My track record includes three sole-authored preprints, a provisional patent on the CCT core architecture, endorsements from Kent Berridge, Samuel Gershman, Nathaniel Daw, and Marcelo Mattar, and functional code repositories for all three platforms. I am an independent researcher based in Lagos, Nigeria, with no institutional affiliation, and I am building these tools because the existing commercial and academic alternatives are closed, expensive, or unavailable in my region.

The requested amount is EUR 35,000 over 12 months. Deliverables include a stable release of TOPOLOGIX v1.0 with documentation and test suite, a containerized GATE v2.0 with web interface, a published software paper, and a workshop for African computational neuroscience researchers on using these tools. All code will remain Apache 2.0.

RESEARCH STATEMENT

My independent research program centers on the Conjunctive Consolidation Threshold (CCT) model, a tripartite pharmacological framework for preventing reward-memory encoding in addiction. The model specifies that three concurrent pharmacological actions, dopamine D1 receptor antagonism, NMDA receptor partial agonism, and mu-opioid receptor antagonism, must exceed a conjunctive threshold to prevent the consolidation of reward-associated memories. I have validated this model through ODE/RK45 simulations and Bayesian MCMC population dynamics, achieving an encoding probability reduction from 0.855 to 0.122 (85.8% reduction) and demonstrating super-additivity of 12.8 percentage points beyond the sum of individual effects. All five pre-registered hypotheses (H1-H5) were confirmed. The formal mathematical specification is published on OSF (10.17605/OSF.IO/EMY4U), and the Bayesian clinical trial architecture is on Zenodo (10.5281/zenodo.20492472). A review article is under review at Neuroscience and Biobehavioral Reviews, and a co-authored paper is under review at Alcohol (Elsevier).

The CCT model is not directly fundable by NLnet. However, the computational tools I built to develop and test the model are precisely the kind of open-source infrastructure NLnet supports. IMPRINT screens compounds for addiction liability using the CCT framework. TOPOLOGIX uses topological data analysis to predict drug-protein interactions and cardiotoxicity. GATE evaluates BCI neural-stimulation safety. All three are Apache 2.0, modular, and designed for reproducibility. The NLnet grant would allow me to generalize TOPOLOGIX beyond hERG cardiotoxicity into a full TDA library for drug safety screening, and to harden GATE into a reusable framework with containerized deployment and a web interface.

The broader impact is that any addiction researcher, pharmacologist, or BCI engineer, especially those in Africa, South Asia, or Latin America, can use these tools without paying for licenses or needing high-performance computing clusters. I am building the infrastructure I needed but did not have when I started this work as an independent researcher in Lagos.

BUDGET NARRATIVE

Total request: EUR 35,000 over 12 months.

Personnel: EUR 20,000 for 12 months of full-time development work by the applicant (Eniola Ayodele Olutogun). This is a subsistence-level stipend for an independent researcher in Lagos, Nigeria, covering rent, food, internet, and electricity. No other personnel are requested.

Computing: EUR 5,000 for cloud compute credits (AWS or Hetzner) for running TDA pipelines, Bayesian MCMC sampling, and container builds. This covers approximately 10,000 CPU-hours and 500 GPU-hours.

Software dependencies and licensing: EUR 2,000 for DOI registration, software deposit fees (Zenodo, Software Heritage), and any required licenses for third-party libraries that are not fully open-source.

Workshop and dissemination: EUR 5,000 for hosting a two-day virtual workshop for African computational neuroscience researchers, including speaker honoraria (EUR 500 each for three invited speakers), platform costs, and recording/transcription. Remaining EUR 3,000 for travel to one international conference (e.g., Computational and Systems Neuroscience, COSYNE) to present the tools and recruit beta testers.

Contingency: EUR 3,000 (approximately 8.6%) for unforeseen costs, exchange rate fluctuations, or extended compute time.

All funds will be managed through a personal account in Nigeria. No institutional overhead is charged.

TIMELINE AND MILESTONES

Month 1-2: Audit existing codebases for TOPOLOGIX and GATE. Write unit tests for all core functions. Publish current versions on Zenodo with DOIs. Begin documentation in Sphinx.

Month 3-4: Implement standardized API for TOPOLOGIX. Add support for user-supplied molecular descriptors and protein structures. Write tutorial notebooks for three use cases: hERG cardiotoxicity screening, dopamine receptor binding prediction, and opioid receptor selectivity analysis.

Month 5-6: Containerize GATE with Docker. Add web-based configuration interface using Flask or FastAPI. Publish benchmark dataset of simulated BCI stimulation outcomes (100 parameter combinations, 10,000 simulated trials).

Month 7-8: Beta release of TOPOLOGIX v0.9 and GATE v2.0. Recruit 5-10 beta testers from the computational neuroscience community (contacts via Berridge, Gershman, Daw, Mattar labs). Collect feedback and fix bugs.

Month 9-10: Final release of TOPOLOGIX v1.0 and GATE v2.0. Write and submit software paper to the Journal of Open Source Software (JOSS) or similar. Prepare workshop materials.

Month 11: Host two-day virtual workshop for African computational neuroscience researchers. Record all sessions, publish on YouTube under CC-BY license.

Month 12: Final report to NLnet. Deposit all code, documentation, and workshop materials on Zenodo with persistent identifiers. Publish blog post summarizing project outcomes.

CHECKLIST

- [ ] Motivation letter (500 words, written above)
- [ ] Research statement (400-600 words, written above)
- [ ] Budget narrative (written above)
- [ ] Timeline and milestones (written above)
- [ ] CV or resume (applicant must prepare from profile data)
- [ ] Links to GitHub repositories (github.com/AmunRaPtah/IMPRINT, github.com/AmunRaPtah/TOPOLOGIX, github.com/AmunRaPtah/GATE)
- [ ] Links to preprints (OSF 10.17605/OSF.IO/KG7B5, OSF 10.17605/OSF.IO/EMY4U, Zenodo 10.5281/zenodo.20492472)
- [ ] ORCID profile (0009-0001-9272-6735)
- [ ] Proof of open-source licensing (Apache 2.0 for all three platforms)
- [ ] Two letters of recommendation (suggested: Kent Berridge, Samuel Gershman)
- [ ] Provisional patent documentation (CCT core architecture, Q3 2026)
- [ ] Identification document (passport or national ID)
- [ ] Bank account details for fund transfer (Nigerian bank account)

EDITOR NOTES

- Eligibility risk: NLnet typically funds organizations or individuals with a bank account in the EU/EEA. The applicant is based in Nigeria. Confirm whether NLnet accepts applications from Nigerian individuals, or whether a fiscal sponsor is needed. If not, consider applying through a European collaborator or open-source foundation (e.g., Software Freedom Conservancy).
- Verification needed: Confirm that the three platforms (IMPRINT, TOPOLOGIX, GATE) are indeed licensed under Apache 2.0 and that the GitHub repositories are public and contain functional code. The profile mentions them but does not include repository URLs. The applicant must provide these.
- Gap: The profile does not specify the exact amount requested. EUR 35,000 is an estimate based on typical NLnet grant size and the budget narrative. The applicant should confirm this figure aligns with their actual needs and NLnet's typical range (EUR 5,000-50,000).
- Gap: The profile does not mention any prior experience managing grants or budgets. The applicant should be prepared to explain how they will handle financial reporting and accountability as an independent researcher.
- Gap: The workshop for African computational neuroscience researchers is mentioned in the timeline but not detailed in the profile. The applicant should confirm they have contacts or a mailing list to recruit participants, or plan to partner with an existing organization (e.g., African Institute for Mathematical Sciences, or the Society for Neuroscientists of Africa).

AI Draft — NLnet Foundation Grant