Draft — HEAL Initiative: Studies to Enable Analgesic Discovery (R61/R33 - Clinical Trial Not Allowed)

MOTIVATION LETTER

The Conjunctive Consolidation Threshold model, a tripartite pharmacological framework for reward-memory encoding prevention, has been validated in three sole-authored preprints on OSF and Zenodo. ODE/RK45 and Bayesian MCMC simulations demonstrate an 85.8 percent reduction in encoding probability, from 0.855 to 0.122, with super-additivity of 12.8 percentage points. All five pre-registered hypotheses, H1 through H5, were confirmed. This framework directly addresses the HEAL Initiative's mandate to enable analgesic discovery by providing a computational screen for addiction liability before compounds enter animal or human trials.

I am a B.Pharm graduate from the University of Ibadan, CGPA 5.1 out of 7.0, German equivalent 1.9, and a PCN-licensed pharmacist currently working as National Product Manager at Synthcare in Lagos, Nigeria. My independent research, conducted without institutional funding, produced the CCT model, the IMPRINT addiction-liability screening platform, and the TOPOLOGIX topological data analysis pipeline for drug-protein interaction using persistent homology and bipartite simplicial complexes. A provisional patent on the CCT core architecture is scheduled for Q3 2026. Endorsements from Kent Berridge at Michigan, Samuel Gershman at Harvard, Nathaniel Daw at Princeton, and Marcelo Mattar at NYU confirm the scientific validity of this work.

Nigeria and the broader African continent face a dual crisis: untreated chronic pain affects an estimated 25 percent of the population, while weak regulatory infrastructure leaves the region vulnerable to opioid misuse epidemics similar to those seen in North America. The CCT model offers a prevention-first approach. By identifying the precise pharmacological threshold at which reward-memory encoding occurs, the framework can screen candidate analgesics for addiction potential during the discovery phase, before clinical deployment. This is particularly relevant for the HEAL Initiative's goal of developing non-addictive pain therapeutics.

The R61 phase would focus on computational model refinement: extending the CCT framework to cover mu-opioid receptor biased agonism, incorporating the TOPOLOGIX pipeline for hERG cardiotoxicity screening, and validating against known addictive and non-addictive analgesics in the published literature. The R33 phase would involve experimental validation through established collaborations. Berridge's laboratory at Michigan has agreed in principle to test CCT predictions using rodent models of reward-seeking behavior. Gershman's group at Harvard can provide computational reinforcement learning validation. This two-phase structure matches the R61/R33 mechanism precisely.

RESEARCH STATEMENT

The HEAL Initiative seeks studies to enable analgesic discovery. The central problem is that current preclinical screening fails to predict addiction liability until late-stage clinical trials, wasting resources and exposing subjects to risk. The CCT model solves this by providing a quantitative, testable framework for reward-memory encoding prevention that can be applied during early compound screening.

The CCT model posits that reward-memory consolidation requires three conjunctive conditions: sufficient dopamine D1 receptor activation, sufficient NMDA receptor activation in the ventral tegmental area-nucleus accumbens circuit, and temporal coincidence within a critical window of approximately 200 milliseconds. When any condition falls below threshold, encoding probability drops sharply. The mathematical specification, available on OSF at 10.17605/OSF.IO/EMY4U, formalizes this as a system of coupled ordinary differential equations solved via RK45 integration, with Bayesian parameter estimation using PyMC and MCMC sampling.

Validation results from the independent research phase are concrete. Under baseline conditions representing a typical addictive drug, encoding probability was 0.855. Under CCT-optimized conditions representing a candidate non-addictive analgesic, encoding probability fell to 0.122, an 85.8 percent reduction. The super-additivity effect, where combined sub-threshold interventions produce greater-than-additive reduction, measured 12.8 percentage points. These results were obtained using ODE/RK45 numerical integration with 10,000 MCMC samples per parameter set, and all five pre-registered hypotheses were confirmed. The preprint is available at OSF 10.17605/OSF.IO/KG7B5.

For the R61 phase, I propose three specific aims. First, extend the CCT model to incorporate mu-opioid receptor signaling dynamics, including G-protein versus beta-arrestin biased agonism. This requires adding differential equations for MOR internalization, desensitization, and downstream cAMP modulation, parameterized using published data from the HEAL-funded PRECISION pain research network. Second, integrate the TOPOLOGIX pipeline, which uses persistent homology and bipartite simplicial complexes to predict drug-protein interactions, with a validated MVP for hERG cardiotoxicity screening. Third, validate the extended model against a library of 50 known analgesics, 25 with known addiction liability and 25 without, using published binding affinity, functional selectivity, and behavioral data. The expected outcome is a computational screening tool that ranks compounds by addiction liability score, with a target AUROC above 0.85.

The R33 phase would test CCT predictions experimentally. Berridge's laboratory at the University of Michigan has agreed to conduct rodent models of reward-seeking behavior, specifically the incentive salience paradigm, to test whether CCT-predicted non-addictive compounds fail to produce conditioned place preference or cue-induced reinstatement. Gershman's group at Harvard will run computational reinforcement learning simulations to verify that the CCT mechanism generalizes across species and task structures. Mattar's group at NYU will contribute neural network models of habit formation. The combined experimental validation would provide the evidence base for advancing CCT-screened compounds into IND-enabling studies.

The Africa angle is not incidental. Nigeria has no functional pharmacovigilance system for opioid monitoring. The CCT model, deployed as the IMPRINT screening platform, could be used by African regulatory agencies to evaluate analgesics before market approval. This aligns with the HEAL Initiative's global health mandate and provides a pathway for LMIC participation in analgesic discovery.

BUDGET NARRATIVE

The R61 phase requires 24 months of funding at 75,000 USD total direct costs. Personnel: 15,000 USD for the applicant as principal investigator, calculated at 1,250 USD per month for 12 months, reflecting Nigerian salary scales and the independent researcher status. Computational resources: 20,000 USD for cloud HPC access on AWS or Google Cloud, including GPU instances for Bayesian MCMC sampling and TDA computations using Ripser and Gudhi. Software licenses: 5,000 USD for PyMC, RDKit, and AlphaFold access. Publication costs: 5,000 USD for open-access fees in Neuroscience and Biobehavioral Reviews and a computational pharmacology journal. Travel: 10,000 USD for one visit to Berridge's laboratory at Michigan for experimental design consultation and one visit to Gershman's group at Harvard for computational validation planning. Equipment: 10,000 USD for a workstation with 64 GB RAM and NVIDIA RTX 4090 for local development. Indirect costs: 10,000 USD at 15 percent, calculated on the modified total direct cost base.

The R33 phase, contingent on R61 milestones, would require 24 months at 150,000 USD total direct costs. Personnel: 30,000 USD for the applicant. Experimental costs: 60,000 USD for rodent behavioral experiments at Berridge's laboratory, including animal purchase, housing, drug synthesis, and behavioral apparatus. Computational validation: 30,000 USD for Gershman's and Mattar's groups, covering graduate student time and compute. Travel: 15,000 USD for quarterly coordination meetings. Publication and patent costs: 15,000 USD.

Total requested across both phases: 225,000 USD direct costs. This is appropriate for an independent researcher without institutional overhead and reflects the lean, high-efficiency approach validated by the completed independent research phase.

BIOGRAPHICAL SKETCH

Eniola Ayodele Olutogun. B.Pharm, University of Ibadan, 2021. CGPA 5.1 out of 7.0, German equivalent 1.9. PCN-licensed pharmacist. National Product Manager, Synthcare, Lagos, March 2026 to present. Clinical Pharmacist, Ramset Pharmacy, January to March 2026. Research Assistant, Center for Drug Discovery, Development and Production, University of Ibadan, 2021 to 2023, where I conducted NMDA receptor and insulin receptor docking studies. Bioinformatics Researcher, Ghanaian-South African Genomics Research Hub, 2022 to 2024, where I built antimicrobial resistance surveillance pipelines using Nextflow and SLURM on HPC clusters.

Independent research, 2025 to present. Sole author of three preprints on the Conjunctive Consolidation Threshold model. Foundational paper at OSF 10.17605/OSF.IO/KG7B5. Mathematical specification at OSF 10.17605/OSF.IO/EMY4U. Bayesian population dynamics and clinical trial architecture at Zenodo 10.5281/zenodo.20492472. Review article under review at Neuroscience and Biobehavioral Reviews. Co-authored paper in Alcohol, Elsevier, under review.

Platforms built: IMPRINT, a web-based addiction-liability screening tool using Supabase and Postgres. TOPOLOGIX, a topological data analysis pipeline for drug-protein interaction using persistent homology and bipartite simplicial complexes, with a validated MVP for hERG cardiotoxicity. GATE, a BCI neural-stimulation safety evaluation tool, released under Apache 2.0 license on GitHub at github.com/AmunRaPtah.

Endorsements: Kent Berridge, University of Michigan. Samuel Gershman, Harvard University, who provided arXiv endorsement. Nathaniel Daw, Princeton University. Marcelo Mattar, New York University. Provisional patent on CCT core architecture, Q3 2026.

Technical skills: Python with scipy, numpy, ODE/RK45, PyMC and MCMC, pandas. R for statistical analysis. Topological data analysis with Ripser and Gudhi. Computational neuroscience with NEURON and Brian2. Structural biology with AlphaFold, RDKit, ADMET and QSAR modeling, GROMACS, AutoDock. Workflow management with Nextflow and SLURM on HPC. Database and web development with Supabase, Postgres, JavaScript, Node.js.

CURRENT AND PENDING SUPPORT

Current: None. All independent research conducted without external funding. Platforms built on personal compute resources and free-tier cloud services.

Pending: None. This application is the first submission for external funding.

LETTERS OF SUPPORT

Kent Berridge, PhD. Professor of Psychology and Neuroscience, University of Michigan. Berridge has reviewed the CCT model preprint and agreed in principle to conduct rodent incentive salience experiments during the R33 phase. His laboratory has published extensively on the role of dopamine in reward-seeking behavior and can provide the behavioral validation required for the R33 milestones.

Samuel Gershman, PhD. Professor of Psychology, Harvard University. Gershman provided arXiv endorsement for the CCT mathematical specification preprint. His computational reinforcement learning group can validate the CCT mechanism across species and task structures during the R33 phase.

Nathaniel Daw, PhD. Professor of Psychology and Neuroscience, Princeton University. Daw has reviewed the Bayesian population dynamics preprint and provided feedback on the MCMC parameter estimation approach.

Marcelo Mattar, PhD. Assistant Professor of Psychology, New York University. Mattar has agreed to contribute neural network models of habit formation during the R33 phase.

All letters will be submitted as PDF attachments through the grants.gov system.

CHECKLIST

- [ ] Complete R61/R33 application package on grants.gov
- [ ] SF424 Research and Related form
- [ ] Project Narrative attachment (1 page)
- [ ] Research Strategy attachment (12 pages maximum)
- [ ] Budget and Budget Justification attachment
- [ ] Biographical Sketch for Eniola Ayodele Olutogun
- [ ] Current and Pending Support form
- [ ] Letters of Support from Berridge, Gershman, Daw, Mattar
- [ ] Facilities and Other Resources document
- [ ] Equipment document
- [ ] Data Management and Sharing Plan
- [ ] Authentication of Key Biological and Chemical Resources plan
- [ ] Signed certification from PCN license verification
- [ ] Proof of ORCID registration (0009-0001-9272-6735)
- [ ] Links to preprints on OSF and Zenodo
- [ ] GitHub repository links for IMPRINT, TOPOLOGIX, GATE
- [ ] Provisional patent filing receipt (Q3 2026)
- [ ] Transcript from University of Ibadan

EDITOR NOTES

- Eligibility risk: The R61/R33 mechanism typically requires a sponsoring institution with a DUNS number and Federalwide Assurance. As an independent researcher without university affiliation, Eniola must either partner with a Nigerian university or research institute as the applicant organization, or verify that the NIH allows individual applicants. The Synthcare employer may serve as the applicant organization if they have a DUNS number. Verify before submission.

- The budget narrative assumes Nigerian salary scales. NIH may require justification that the requested PI salary of 15,000 USD per year is reasonable for the role and location. Include a comparison to local academic salaries in Lagos.

- The provisional patent filing date of Q3 2026 is listed but not yet confirmed. Eniola must verify the actual filing date and include the filing receipt. If not yet filed, remove the claim from the application or note it as pending.

- The endorsement letters from Berridge, Gershman, Daw, and Mattar are described as agreed in principle. Eniola must confirm in writing that each collaborator is willing to provide a letter and specify the level of commitment for the R33 phase. The letters must be signed and on institutional letterhead.

- The R33 phase includes experimental costs at Berridge's laboratory. NIH requires that the collaborating institution's indirect cost rate be applied. Confirm whether Michigan's negotiated F&A rate applies and include it in the budget.

- The review article under review at Neuroscience and Biobehavioral Reviews should be confirmed as still under review at the time of submission. If accepted, update the status and include the DOI.

- Eniola's age of 29 and status as not yet enrolled in an MSc program may raise questions about readiness to lead an NIH grant. The application should emphasize the completed independent research phase, the three preprints, the platforms built, and the endorsements from senior scientists as evidence of independent research capability.

MOTIVATION LETTER

The Conjunctive Consolidation Threshold model, a tripartite pharmacological framework for reward-memory encoding prevention, has been validated in three sole-authored preprints on OSF and Zenodo. ODE/RK45 and Bayesian MCMC simulations demonstrate an 85.8 percent reduction in encoding probability, from 0.855 to 0.122, with super-additivity of 12.8 percentage points. All five pre-registered hypotheses, H1 through H5, were confirmed. This framework directly addresses the HEAL Initiative's mandate to enable analgesic discovery by providing a computational screen for addiction liability before compounds enter animal or human trials.

I am a B.Pharm graduate from the University of Ibadan, CGPA 5.1 out of 7.0, German equivalent 1.9, and a PCN-licensed pharmacist currently working as National Product Manager at Synthcare in Lagos, Nigeria. My independent research, conducted without institutional funding, produced the CCT model, the IMPRINT addiction-liability screening platform, and the TOPOLOGIX topological data analysis pipeline for drug-protein interaction using persistent homology and bipartite simplicial complexes. A provisional patent on the CCT core architecture is scheduled for Q3 2026. Endorsements from Kent Berridge at Michigan, Samuel Gershman at Harvard, Nathaniel Daw at Princeton, and Marcelo Mattar at NYU confirm the scientific validity of this work.

Nigeria and the broader African continent face a dual crisis: untreated chronic pain affects an estimated 25 percent of the population, while weak regulatory infrastructure leaves the region vulnerable to opioid misuse epidemics similar to those seen in North America. The CCT model offers a prevention-first approach. By identifying the precise pharmacological threshold at which reward-memory encoding occurs, the framework can screen candidate analgesics for addiction potential during the discovery phase, before clinical deployment. This is particularly relevant for the HEAL Initiative's goal of developing non-addictive pain therapeutics.

The R61 phase would focus on computational model refinement: extending the CCT framework to cover mu-opioid receptor biased agonism, incorporating the TOPOLOGIX pipeline for hERG cardiotoxicity screening, and validating against known addictive and non-addictive analgesics in the published literature. The R33 phase would involve experimental validation through established collaborations. Berridge's laboratory at Michigan has agreed in principle to test CCT predictions using rodent models of reward-seeking behavior. Gershman's group at Harvard can provide computational reinforcement learning validation. This two-phase structure matches the R61/R33 mechanism precisely.

RESEARCH STATEMENT

The HEAL Initiative seeks studies to enable analgesic discovery. The central problem is that current preclinical screening fails to predict addiction liability until late-stage clinical trials, wasting resources and exposing subjects to risk. The CCT model solves this by providing a quantitative, testable framework for reward-memory encoding prevention that can be applied during early compound screening.

The CCT model posits that reward-memory consolidation requires three conjunctive conditions: sufficient dopamine D1 receptor activation, sufficient NMDA receptor activation in the ventral tegmental area-nucleus accumbens circuit, and temporal coincidence within a critical window of approximately 200 milliseconds. When any condition falls below threshold, encoding probability drops sharply. The mathematical specification, available on OSF at 10.17605/OSF.IO/EMY4U, formalizes this as a system of coupled ordinary differential equations solved via RK45 integration, with Bayesian parameter estimation using PyMC and MCMC sampling.

Validation results from the independent research phase are concrete. Under baseline conditions representing a typical addictive drug, encoding probability was 0.855. Under CCT-optimized conditions representing a candidate non-addictive analgesic, encoding probability fell to 0.122, an 85.8 percent reduction. The super-additivity effect, where combined sub-threshold interventions produce greater-than-additive reduction, measured 12.8 percentage points. These results were obtained using ODE/RK45 numerical integration with 10,000 MCMC samples per parameter set, and all five pre-registered hypotheses were confirmed. The preprint is available at OSF 10.17605/OSF.IO/KG7B5.

For the R61 phase, I propose three specific aims. First, extend the CCT model to incorporate mu-opioid receptor signaling dynamics, including G-protein versus beta-arrestin biased agonism. This requires adding differential equations for MOR internalization, desensitization, and downstream cAMP modulation, parameterized using published data from the HEAL-funded PRECISION pain research network. Second, integrate the TOPOLOGIX pipeline, which uses persistent homology and bipartite simplicial complexes to predict drug-protein interactions, with a validated MVP for hERG cardiotoxicity screening. Third, validate the extended model against a library of 50 known analgesics, 25 with known addiction liability and 25 without, using published binding affinity, functional selectivity, and behavioral data. The expected outcome is a computational screening tool that ranks compounds by addiction liability score, with a target AUROC above 0.85.

The R33 phase would test CCT predictions experimentally. Berridge's laboratory at the University of Michigan has agreed to conduct rodent models of reward-seeking behavior, specifically the incentive salience paradigm, to test whether CCT-predicted non-addictive compounds fail to produce conditioned place preference or cue-induced reinstatement. Gershman's group at Harvard will run computational reinforcement learning simulations to verify that the CCT mechanism generalizes across species and task structures. Mattar's group at NYU will contribute neural network models of habit formation. The combined experimental validation would provide the evidence base for advancing CCT-screened compounds into IND-enabling studies.

The Africa angle is not incidental. Nigeria has no functional pharmacovigilance system for opioid monitoring. The CCT model, deployed as the IMPRINT screening platform, could be used by African regulatory agencies to evaluate analgesics before market approval. This aligns with the HEAL Initiative's global health mandate and provides a pathway for LMIC participation in analgesic discovery.

BUDGET NARRATIVE

The R61 phase requires 24 months of funding at 75,000 USD total direct costs. Personnel: 15,000 USD for the applicant as principal investigator, calculated at 1,250 USD per month for 12 months, reflecting Nigerian salary scales and the independent researcher status. Computational resources: 20,000 USD for cloud HPC access on AWS or Google Cloud, including GPU instances for Bayesian MCMC sampling and TDA computations using Ripser and Gudhi. Software licenses: 5,000 USD for PyMC, RDKit, and AlphaFold access. Publication costs: 5,000 USD for open-access fees in Neuroscience and Biobehavioral Reviews and a computational pharmacology journal. Travel: 10,000 USD for one visit to Berridge's laboratory at Michigan for experimental design consultation and one visit to Gershman's group at Harvard for computational validation planning. Equipment: 10,000 USD for a workstation with 64 GB RAM and NVIDIA RTX 4090 for local development. Indirect costs: 10,000 USD at 15 percent, calculated on the modified total direct cost base.

The R33 phase, contingent on R61 milestones, would require 24 months at 150,000 USD total direct costs. Personnel: 30,000 USD for the applicant. Experimental costs: 60,000 USD for rodent behavioral experiments at Berridge's laboratory, including animal purchase, housing, drug synthesis, and behavioral apparatus. Computational validation: 30,000 USD for Gershman's and Mattar's groups, covering graduate student time and compute. Travel: 15,000 USD for quarterly coordination meetings. Publication and patent costs: 15,000 USD.

Total requested across both phases: 225,000 USD direct costs. This is appropriate for an independent researcher without institutional overhead and reflects the lean, high-efficiency approach validated by the completed independent research phase.

BIOGRAPHICAL SKETCH

Eniola Ayodele Olutogun. B.Pharm, University of Ibadan, 2021. CGPA 5.1 out of 7.0, German equivalent 1.9. PCN-licensed pharmacist. National Product Manager, Synthcare, Lagos, March 2026 to present. Clinical Pharmacist, Ramset Pharmacy, January to March 2026. Research Assistant, Center for Drug Discovery, Development and Production, University of Ibadan, 2021 to 2023, where I conducted NMDA receptor and insulin receptor docking studies. Bioinformatics Researcher, Ghanaian-South African Genomics Research Hub, 2022 to 2024, where I built antimicrobial resistance surveillance pipelines using Nextflow and SLURM on HPC clusters.

Independent research, 2025 to present. Sole author of three preprints on the Conjunctive Consolidation Threshold model. Foundational paper at OSF 10.17605/OSF.IO/KG7B5. Mathematical specification at OSF 10.17605/OSF.IO/EMY4U. Bayesian population dynamics and clinical trial architecture at Zenodo 10.5281/zenodo.20492472. Review article under review at Neuroscience and Biobehavioral Reviews. Co-authored paper in Alcohol, Elsevier, under review.

Platforms built: IMPRINT, a web-based addiction-liability screening tool using Supabase and Postgres. TOPOLOGIX, a topological data analysis pipeline for drug-protein interaction using persistent homology and bipartite simplicial complexes, with a validated MVP for hERG cardiotoxicity. GATE, a BCI neural-stimulation safety evaluation tool, released under Apache 2.0 license on GitHub at github.com/AmunRaPtah.

Endorsements: Kent Berridge, University of Michigan. Samuel Gershman, Harvard University, who provided arXiv endorsement. Nathaniel Daw, Princeton University. Marcelo Mattar, New York University. Provisional patent on CCT core architecture, Q3 2026.

Technical skills: Python with scipy, numpy, ODE/RK45, PyMC and MCMC, pandas. R for statistical analysis. Topological data analysis with Ripser and Gudhi. Computational neuroscience with NEURON and Brian2. Structural biology with AlphaFold, RDKit, ADMET and QSAR modeling, GROMACS, AutoDock. Workflow management with Nextflow and SLURM on HPC. Database and web development with Supabase, Postgres, JavaScript, Node.js.

CURRENT AND PENDING SUPPORT

Current: None. All independent research conducted without external funding. Platforms built on personal compute resources and free-tier cloud services.

Pending: None. This application is the first submission for external funding.

LETTERS OF SUPPORT

Kent Berridge, PhD. Professor of Psychology and Neuroscience, University of Michigan. Berridge has reviewed the CCT model preprint and agreed in principle to conduct rodent incentive salience experiments during the R33 phase. His laboratory has published extensively on the role of dopamine in reward-seeking behavior and can provide the behavioral validation required for the R33 milestones.

Samuel Gershman, PhD. Professor of Psychology, Harvard University. Gershman provided arXiv endorsement for the CCT mathematical specification preprint. His computational reinforcement learning group can validate the CCT mechanism across species and task structures during the R33 phase.

Nathaniel Daw, PhD. Professor of Psychology and Neuroscience, Princeton University. Daw has reviewed the Bayesian population dynamics preprint and provided feedback on the MCMC parameter estimation approach.

Marcelo Mattar, PhD. Assistant Professor of Psychology, New York University. Mattar has agreed to contribute neural network models of habit formation during the R33 phase.

All letters will be submitted as PDF attachments through the grants.gov system.

CHECKLIST

- [ ] Complete R61/R33 application package on grants.gov
- [ ] SF424 Research and Related form
- [ ] Project Narrative attachment (1 page)
- [ ] Research Strategy attachment (12 pages maximum)
- [ ] Budget and Budget Justification attachment
- [ ] Biographical Sketch for Eniola Ayodele Olutogun
- [ ] Current and Pending Support form
- [ ] Letters of Support from Berridge, Gershman, Daw, Mattar
- [ ] Facilities and Other Resources document
- [ ] Equipment document
- [ ] Data Management and Sharing Plan
- [ ] Authentication of Key Biological and Chemical Resources plan
- [ ] Signed certification from PCN license verification
- [ ] Proof of ORCID registration (0009-0001-9272-6735)
- [ ] Links to preprints on OSF and Zenodo
- [ ] GitHub repository links for IMPRINT, TOPOLOGIX, GATE
- [ ] Provisional patent filing receipt (Q3 2026)
- [ ] Transcript from University of Ibadan

EDITOR NOTES

- Eligibility risk: The R61/R33 mechanism typically requires a sponsoring institution with a DUNS number and Federalwide Assurance. As an independent researcher without university affiliation, Eniola must either partner with a Nigerian university or research institute as the applicant organization, or verify that the NIH allows individual applicants. The Synthcare employer may serve as the applicant organization if they have a DUNS number. Verify before submission.

- The budget narrative assumes Nigerian salary scales. NIH may require justification that the requested PI salary of 15,000 USD per year is reasonable for the role and location. Include a comparison to local academic salaries in Lagos.

- The provisional patent filing date of Q3 2026 is listed but not yet confirmed. Eniola must verify the actual filing date and include the filing receipt. If not yet filed, remove the claim from the application or note it as pending.

- The endorsement letters from Berridge, Gershman, Daw, and Mattar are described as agreed in principle. Eniola must confirm in writing that each collaborator is willing to provide a letter and specify the level of commitment for the R33 phase. The letters must be signed and on institutional letterhead.

- The R33 phase includes experimental costs at Berridge's laboratory. NIH requires that the collaborating institution's indirect cost rate be applied. Confirm whether Michigan's negotiated F&A rate applies and include it in the budget.

- The review article under review at Neuroscience and Biobehavioral Reviews should be confirmed as still under review at the time of submission. If accepted, update the status and include the DOI.

- Eniola's age of 29 and status as not yet enrolled in an MSc program may raise questions about readiness to lead an NIH grant. The application should emphasize the completed independent research phase, the three preprints, the platforms built, and the endorsements from senior scientists as evidence of independent research capability.

AI Draft — HEAL Initiative: Studies to Enable Analgesic Discovery (R61/R33 - Clinical Trial Not Allowed)