MOTIVATION LETTER The Conjunctive Consolidation Threshold model, or CCT, is a tripartite pharmacological framework that prevents reward-memory encoding in addiction. I developed this model as an independent researcher in Lagos, Nigeria, and validated it through five pre-registered hypotheses using ODE/RK45 and Bayesian MCMC methods. The model reduces encoding probability from 0.855 to 0.122, an 85.8 percent reduction, with super-additivity of 12.8 percentage points. Three sole-authored preprints on OSF and Zenodo document the foundational theory, formal mathematical specification, and Bayesian population dynamics with clinical trial architecture. A review article is under review at Neuroscience and Biobehavioral Reviews, and a co-authored paper is under review at Alcohol. The NSF Small Business Innovation Research and Small Business Technology Transfer program, with its pilot emphasis on scientific instrumentation, is the correct vehicle to translate the CCT model into a deployable computational platform for addiction neuroscience. I have built three platforms that serve as the instrumentation layer for this translation. IMPRINT screens addiction liability using Bayesian inference. TOPOLOGIX applies topological data analysis, persistent homology, and bipartite simplicial complexes to drug-protein interaction, with a validated hERG cardiotoxicity MVP. GATE evaluates BCI neural-stimulation safety under Apache 2.0. These platforms, combined with the CCT model, constitute a next-generation, AI-driven scientific instrumentation suite that enables high-throughput, pre-registered screening of reward-memory mechanisms. The NSF mission to strengthen economic growth and improve lives through scientific discovery aligns directly with this work. Addiction imposes a measurable burden on the U.S. healthcare system and on global public health. A computational platform that predicts addiction liability before clinical deployment of a compound would accelerate U.S. research on reward-memory mechanisms and reduce the failure rate of addiction therapeutics in late-stage trials. The gold-standard validation of the CCT model, including Bayesian MCMC confirmation of all five pre-registered hypotheses, meets the NSF standard for rigorous, reproducible science. I hold a B.Pharm from the University of Ibadan with a CGPA of 5.1 out of 7.0, a German equivalent of 1.9, and I am a PCN-licensed pharmacist. I have endorsements from Kent Berridge at the University of Michigan, Samuel Gershman at Harvard, Nathaniel Daw at Princeton, and Marcelo Mattar at NYU. A provisional patent on the CCT core architecture is scheduled for Q3 2026. I am currently National Product Manager at Synthcare in Lagos and an independent researcher affiliated with ZYCO. The eligibility requirement for a U.S.-registered small business is the central structural challenge. I am prepared to establish a U.S. subsidiary of ZYCO or to partner with a U.S.-based small business that specializes in computational neuroscience instrumentation. I request guidance from the NSF program officer on the preferred pathway for an independent researcher based in Nigeria. RESEARCH STATEMENT The Conjunctive Consolidation Threshold model addresses a specific gap in addiction neuroscience: the absence of a formal, mathematically specified framework that predicts the encoding of reward-memory associations at the pharmacological level. Current models describe addiction as a disorder of dopamine signaling or habit formation, but they do not specify the threshold conditions under which a reward event becomes consolidated into a persistent memory trace. The CCT model fills this gap by defining a tripartite interaction among dopamine D1 receptor activation, NMDA receptor-mediated calcium influx, and cAMP response element-binding protein phosphorylation. When these three signals exceed a conjunctive threshold within a defined temporal window, reward-memory encoding occurs. Below that threshold, encoding is prevented. The formal mathematical specification, available on OSF at DOI 10.17605/OSF.IO/EMY4U, expresses the CCT as a system of ordinary differential equations solved with RK45 integration. The Bayesian population dynamics model, on Zenodo at DOI 10.5281/zenodo.20492472, uses Markov Chain Monte Carlo sampling to estimate parameter distributions across a simulated population. The results confirm all five pre-registered hypotheses. Encoding probability drops from 0.855 to 0.122 under combined pharmacological intervention. Super-additivity of 12.8 percentage points indicates that the three components act synergistically, not additively. The three platforms I built operationalize this model for experimental and clinical use. IMPRINT is a Bayesian screening tool that takes a compound's receptor binding profile and outputs a predicted addiction liability score with credible intervals. TOPOLOGIX uses persistent homology and bipartite simplicial complexes to analyze drug-protein interaction networks, with a validated MVP for hERG cardiotoxicity prediction. GATE is a BCI neural-stimulation safety evaluation tool released under Apache 2.0. Together, these platforms form an integrated instrumentation suite for addiction neuroscience. The NSF SBIR/STTR program with its pilot emphasis on scientific instrumentation is the appropriate funding mechanism because the CCT suite is not a drug or a therapy. It is a computational instrument that enables researchers to screen compounds, predict liability, and design experiments with pre-registered, falsifiable hypotheses. The program's focus on AI-enabled discovery tools matches the Bayesian and topological data analysis methods embedded in the platform. The Phase I objective would be to build a minimum viable product of the integrated suite, validated against a benchmark set of 50 known addictive and non-addictive compounds. Phase II would expand to a cloud-deployed platform with a graphical user interface, API access, and a validation study with a U.S. academic partner. The provisional patent filing in Q3 2026 covers the core CCT architecture, including the conjunctive threshold algorithm and the Bayesian inference engine for population-level prediction. This intellectual property provides the basis for a commercializable instrument. The endorsements from Berridge, Gershman, Daw, and Mattar confirm that the theoretical foundation is sound and that the platform addresses a recognized need in the field. PROJECT DESCRIPTION The proposed project, titled CCT-Instrument, aims to develop and validate an integrated computational platform for addiction liability screening. The platform combines the Conjunctive Consolidation Threshold model with three existing software tools: IMPRINT, TOPOLOGIX, and GATE. The objective is to produce a scientific instrument that researchers can use to predict whether a novel compound will produce reward-memory encoding, and therefore addiction liability, before the compound enters animal or human trials. The specific aims are three. First, integrate the CCT Bayesian inference engine with IMPRINT's screening pipeline to produce a unified prediction tool that accepts molecular input data and outputs addiction liability scores with uncertainty quantification. Second, incorporate TOPOLOGIX's topological data analysis module to identify structural features in drug-protein interaction networks that correlate with CCT threshold crossing. Third, deploy GATE as a safety evaluation layer for compounds that pass the initial screen, specifically for compounds intended for BCI-based neuromodulation. The technical approach uses the existing codebase. IMPRINT is written in Python with PyMC for Bayesian inference. TOPOLOGIX uses Ripser and Gudhi for persistent homology. GATE is a JavaScript and Node.js application with a Supabase backend. Integration will require building a common data schema, a REST API layer, and a command-line interface for batch processing. Validation will use a benchmark set of 50 compounds with known addiction liability, drawn from the literature and from the Drug Enforcement Administration scheduling database. The primary metric is area under the receiver operating characteristic curve, with a target of 0.85 or higher. The timeline is 12 months for Phase I. Months one through three focus on data schema design and API development. Months four through six focus on integration of the three platforms and the CCT inference engine. Months seven through nine focus on validation against the benchmark set. Months ten through twelve focus on documentation, user testing with two academic collaborators, and preparation of a Phase II proposal. The team consists of myself as principal investigator, with computational pharmacology expertise and software development skills. I will recruit a U.S.-based co-investigator with expertise in addiction neuroscience and animal models to serve as the validation partner. The budget for Phase I is 100,000 U.S. dollars, allocated to software development, cloud computing costs, the co-investigator's consulting fee, and travel for a validation workshop at the partner institution. BUDGET JUSTIFICATION The Phase I budget of 100,000 U.S. dollars supports 12 months of development and validation. Personnel costs total 60,000 dollars. My salary as principal investigator is 40,000 dollars, calculated at 50 percent effort for 12 months at a rate of 80,000 dollars per year. The U.S.-based co-investigator receives a consulting fee of 20,000 dollars for 200 hours of work at 100 dollars per hour, covering experimental design, compound selection, and validation oversight. Equipment and computing costs total 20,000 dollars. Cloud computing on AWS or Google Cloud for Bayesian MCMC sampling and topological data analysis is estimated at 15,000 dollars. A dedicated workstation with a GPU for local development and testing costs 5,000 dollars. Materials and supplies total 5,000 dollars. This covers software licenses for RDKit, GROMACS, and AutoDock, as well as database access fees for chemical structure repositories. Travel costs total 10,000 dollars. This funds one validation workshop at the U.S. partner institution, including airfare from Lagos, accommodation for two weeks, and per diem. Indirect costs total 5,000 dollars, calculated at 5 percent of direct costs. The total direct costs are 95,000 dollars, and total indirect costs are 5,000 dollars, for a total of 100,000 dollars. BIOGRAPHICAL SKETCH Eniola Ayodele Olutogun. Independent researcher, Lagos, Nigeria. Affiliated with ZYCO. ORCID 0009-0001-9272-6735. GitHub github.com/AmunRaPtah. Education. B.Pharm, University of Ibadan, 2014 to 2021. CGPA 5.1 out of 7.0, 2:1 Upper Division, German equivalent 1.9. PCN-licensed pharmacist. Research. Independent research from 2025 to present. Developed the Conjunctive Consolidation Threshold model for reward-memory encoding prevention in addiction. Three sole-authored preprints on OSF and Zenodo. Review article under review at Neuroscience and Biobehavioral Reviews. Co-authored paper under review at Alcohol. Built three computational platforms: IMPRINT for addiction liability screening, TOPOLOGIX for topological data analysis of drug-protein interactions, and GATE for BCI neural-stimulation safety evaluation. Provisional patent on CCT core architecture scheduled for Q3 2026. Employment. National Product Manager, Synthcare, Lagos, March 2026 to present. Clinical Pharmacist, Ramset Pharmacy, Lagos, January to March 2026. Research Assistant, Center for Drug Discovery, Development and Production, University of Ibadan, NMDA and insulin docking studies. Bioinformatics Researcher, Ghanaian-Swedish Antimicrobial Resistance Research Group, AMR genomics and surveillance pipeline. Skills. Python with scipy, numpy, ODE/RK45, PyMC for MCMC, and 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, GROMACS, and AutoDock. High-performance computing with Nextflow, SLURM, and HPC. Database and web development with Supabase, Postgres, JavaScript, and Node.js. Endorsements. Kent Berridge, University of Michigan. Samuel Gershman, Harvard University, provided arXiv endorsement. Nathaniel Daw, Princeton University. Marcelo Mattar, New York University. CHECKLIST - [ ] Confirm U.S. small business registration or partnership agreement before submission - [ ] Register with NSF FastLane or Research.gov as an authorized organizational representative - [ ] Complete the Project Summary form, including intellectual merit and broader impacts statements - [ ] Upload the Project Description document, 15 pages maximum, single-spaced, 12-point font - [ ] Upload the Budget Justification document - [ ] Upload the Biographical Sketch for Eniola Ayodele Olutogun, two pages maximum - [ ] Upload the Current and Pending Support form, listing all current and pending funding - [ ] Upload the Facilities, Equipment, and Other Resources document - [ ] Upload the Data Management Plan, two pages maximum - [ ] Upload the Postdoctoral Mentoring Plan, if applicable - [ ] Obtain signed letters of collaboration from the U.S.-based co-investigator and from Kent Berridge, Samuel Gershman, Nathaniel Daw, or Marcelo Mattar - [ ] Obtain a signed letter from the U.S. partner small business confirming the partnership agreement - [ ] Verify that all preprints and Zenodo records are publicly accessible and have valid DOIs - [ ] Verify that the provisional patent application is filed or in process before the submission deadline - [ ] Submit the application by July 27, 2026, at 5:00 PM submitter's local time EDITOR NOTES - Eligibility is the single greatest risk. The NSF SBIR/STTR program requires a U.S.-registered small business. Eniola must either incorporate a U.S. subsidiary of ZYCO or find a U.S. small business partner willing to serve as the applicant organization. This must be confirmed before any other work begins. - The budget justification assumes a U.S.-based co-investigator. The name and institutional affiliation of this person are not in the profile. Eniola must identify and recruit this person, and obtain a signed letter of collaboration. - The benchmark set of 50 compounds is mentioned but not specified. Eniola should compile a list of 50 compounds with known addiction liability, including sources and references, and include it as a supplementary document or in the Project Description. - The provisional patent filing in Q3 2026 is stated as scheduled. Eniola must confirm that the filing will occur before the NSF submission deadline, or that a provisional application is already on file. The patent status affects the commercialization plan. - The profile states that Eniola is not yet enrolled in an MSc program and is applying for an October 2026 start at MUG or Graz, Austria. The NSF SBIR/STTR program does not require an MSc, but the time commitment for a Phase I project while starting a graduate program may be a conflict. Eniola should clarify the timeline and effort allocation in the Project Description or in a cover letter.