Draft — Wellcome Early Career Award

MOTIVATION LETTER

The Conjunctive Consolidation Threshold model, or CCT, 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, into a single system of ordinary differential equations solved with RK45 and calibrated through Bayesian MCMC using PyMC's DEMetropolisZ sampler. The model has fourteen free parameters, all with literature-elicited priors drawn from a systematic 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 currently under review at peer-reviewed journals, and a co-authored paper is under review at Alcohol, an Elsevier journal.

The Wellcome Early Career Award funds early-career researchers to establish independent research programmes. My profile matches that mandate directly. I am an independent researcher, not a postdoc in someone else's laboratory. I have designed, executed, and reported a complete research cycle, from pre-registration through Bayesian calibration to publication, without institutional supervision. The CCT model is my own construction, and it is ready to be the core of a five-year independent programme.

Wellcome's selection criteria emphasize scientific excellence, feasibility, and potential to advance the field. The CCT model has already demonstrated excellence through its pre-registered design and confirmed hypotheses. It is feasible because the computational infrastructure exists and is tested. Its potential to advance the field lies in its capacity to identify novel pharmacological targets for addiction treatment, specifically targets that interrupt the consolidation of reward memories rather than merely dampening reward sensation.

My Nigerian nationality and my enrollment in the M.Sc. Digital Health programme at the Hasso Plattner Institute, University of Potsdam, make me eligible under Wellcome's LMIC track. I am applying as an early-career researcher with no more than three years of postdoctoral experience, which is satisfied by my current status. The award would fund my salary and research costs, with research costs estimated within Wellcome's stated limit of 400,000 pounds.

The CCT model addresses a significant health-related challenge. Addiction is a chronic, relapsing condition, and current pharmacological interventions target acute reward rather than the memory processes that drive craving and relapse. By identifying the threshold at which reward memories consolidate, the CCT model opens a new class of interventions. This is early-stage, proof-of-concept work, exactly what Wellcome's Early Career Award is designed to support.

I am requesting consideration for this award to build the CCT programme into a full independent research line, with the Wellcome Early Career Award as its funding vehicle.

RESEARCH STATEMENT

The CCT programme asks a specific question: can we identify the pharmacological threshold at which reward memories consolidate, and can we intervene at that threshold to prevent addiction? The model treats reward-memory encoding as a conjunctive process, meaning that three separate neural signals must coincide for consolidation to occur. Those signals are dopaminergic reward prediction error, NMDAR-dependent long-term potentiation, and affective contrast, the emotional valence difference between the drug state and the baseline state. If all three cross their individual thresholds simultaneously, the memory consolidates. If any one is suppressed below threshold, consolidation fails.

The model is implemented as a system of coupled ODEs solved with RK45. It has fourteen free parameters, each with a prior distribution elicited from a systematic literature screen of 1,847 records. Calibration was performed with PyMC's DEMetropolisZ sampler, a Markov chain Monte Carlo method suited to high-dimensional posterior exploration. The model was pre-registered with five hypotheses, H1 through H5, before calibration. All five were confirmed. The key result is posterior super-additivity of 13 to 22 percentage points across model versions, meaning the combined effect of suppressing multiple axes exceeds the sum of suppressing each axis alone. This is the finding with direct therapeutic implications: combination therapy targeting two or three axes simultaneously should be substantially more effective than monotherapy.

The five-year programme has three phases. Phase one, months one to eighteen, extends the CCT model from its current single-compartment form to a multi-region form incorporating mesolimbic and mesocortical circuits. This requires integrating the model with the neurocascade simulation engine I have already built, which couples pharmacokinetics to receptor binding to Wilson-Cowan circuit dynamics to behavioral readouts. neurocascade has 62 passing tests and is Bayesian-calibrated for three receptor systems, mu-opioid, D2 dopamine, and GABA-A. The integration will allow the CCT model to make predictions about circuit-level effects of pharmacological interventions, not just molecular-level effects.

Phase two, months nineteen to thirty-six, uses the integrated model to generate a ranked list of candidate drug combinations for preventing reward-memory consolidation. Each candidate will be evaluated in silico for predicted efficacy, predicted side-effect profile, and predicted blood-brain barrier penetration. The ranking will be published as a pre-registered prediction list before any experimental validation begins. This pre-registration is consistent with my established methodology, which has used pre-registration in the CCT model, the cardiotoxicity topology study, and the ergofluids validation pipeline.

Phase three, months thirty-seven to sixty, focuses on dissemination and collaboration. The model and its predictions will be released as open-source software with full documentation. I will seek experimental collaborators, particularly in Nigeria and other LMIC settings, to test the highest-ranked candidates in preclinical models. The goal is to produce a validated, open computational platform that any addiction neuroscience laboratory can use to screen pharmacological interventions for their effect on reward-memory consolidation.

The CCT programme is distinct from my other research lines. The cardiotoxicity topology study and the TOPOLOGIX project address drug resistance and toxicity prediction using persistent homology and protein language models. Those are methodological contributions to computational pharmacology. The CCT programme is a mechanistic contribution to addiction neuroscience. It is the line that best matches Wellcome's mission to support research with the potential to improve health outcomes, because it directly targets the memory processes that drive addiction relapse.

Budget justification: the award would fund my salary for five years, plus research costs estimated at 250,000 pounds. Research costs include high-performance computing time for Bayesian calibration and simulation, approximately 40,000 pounds; open-access publication fees for approximately ten papers, 20,000 pounds; travel to two conferences per year, 30,000 pounds; and a part-time research assistant for the final two years, 60,000 pounds. The remainder covers software infrastructure, data storage, and contingency.

ESSAY: INDEPENDENCE AND RESEARCH IDENTITY

My research identity is defined by a commitment to pre-registration, Bayesian calibration, and honest reporting of negative results. This commitment is documented in my published work. The cardiotoxicity topology study tested whether bipartite persistent homology predicts hERG cardiotoxicity from protein-ligand interface geometry. The result was negative: topological features did not beat a plain descriptor baseline, with AUROC 0.8426 versus 0.8782. I reported this result directly. The interface-topology-for-resistance study found that the same topological constructs carried almost no signal for drug-resistance prediction, with AUROC 0.425 and 0.485 on the Platinum benchmark. I reported this as a ruling-out, not as a failure. The ergofluids project pre-registered a gated validation pipeline. The synthetic-data gates passed. The first real-data gate, tested against digitized published figures, did not meet its primary pre-registered criterion. I reported this directly rather than reframing the result.

This pattern, pre-register, test, report honestly, is the core of my independence. I do not need a supervisor to enforce scientific rigor because I have built it into my workflow. The CCT model is the strongest expression of this identity. It was pre-registered with five hypotheses, calibrated with Bayesian methods, and all five hypotheses were confirmed. The result is not a lucky correlation; it is a structural finding about how three neural signals interact to produce memory consolidation.

Wellcome's Early Career Award asks for evidence of readiness to lead an independent research programme and develop a distinct research identity. My distinct identity is this: I build computational models of biological systems that are pre-registered, Bayesian-calibrated, and honestly reported, and I use those models to make falsifiable predictions that experimentalists can test. The CCT programme is the vehicle for that identity over the next five years.

ESSAY: FIT WITH WELLCOME'S MISSION

Wellcome's mission is to support research that addresses significant health-related challenges. Addiction is one of the most significant health challenges of the current century, with a global burden measured in millions of disability-adjusted life years and a treatment landscape that has not fundamentally changed in decades. Current pharmacotherapies for addiction, such as opioid replacement therapy and nicotine replacement therapy, target the reward system acutely. They do not target the memory processes that drive craving and relapse. The CCT model identifies a specific, mechanistically grounded intervention point: the conjunctive threshold at which reward memories consolidate. If that threshold can be raised pharmacologically, then drug-associated memories will not consolidate, and relapse risk will drop.

This is early-stage, proof-of-concept work. It is not a clinical trial. It is not a large-scale epidemiological study. It is a computational model that generates specific, testable predictions about which pharmacological targets, and which combinations of targets, are most likely to prevent reward-memory consolidation. Wellcome's Early Career Award explicitly supports this kind of work, early-stage research with the potential to open new therapeutic avenues.

The LMIC angle is direct. Nigeria has one of the highest rates of opioid misuse in West Africa, and the treatment infrastructure is thin. A computational model that can screen drug combinations in silico, before expensive and logistically difficult preclinical trials, is particularly valuable in settings where laboratory capacity is limited. The CCT programme includes a specific plan to seek experimental collaborators in Nigeria and other LMIC settings during phase three. This is built into the programme design.

CHECKLIST

- [ ] Confirm eligibility under Wellcome Early Career Award LMIC track, specifically the OECD LMIC list and the exclusion of China and India
- [ ] Verify that enrollment in M.Sc. Digital Health at HPI/Potsdam does not conflict with the "no more than 3 years postdoctoral experience" criterion
- [ ] Obtain and upload all three sole-authored CCT preprints (OSF/Zenodo) as supporting documents
- [ ] Obtain and upload the co-authored Alcohol paper (under review) as supporting document
- [ ] Prepare a two-page CV in Wellcome's required format, including ORCID 0009-0001-9272-6735 and GitHub github.com/AmunRaPtah
- [ ] Prepare a detailed budget breakdown for 250,000 pounds research costs, with line items for HPC time, publication fees, travel, and research assistant salary
- [ ] Draft a data management plan covering open-source release of CCT model code and pre-registration of phase two predictions
- [ ] Draft a project timeline with milestones for phases one, two, and three
- [ ] Identify and contact two referees who can speak to independent research capability, preferably from the named collaborators (Berridge, Gershman, Daw, Mattar)
- [ ] Verify the exact submission portal requirements at grantsby.eu/grants/wellcome-early-career-award-2026, including any additional forms not captured in this draft
- [ ] Confirm the rolling deadline and plan submission for a date that allows at least four weeks for referee letters

EDITOR NOTES

- Eligibility risk: the profile states enrollment in M.Sc. Digital Health at HPI/Potsdam for Winter Semester 2026/27, but the employment section lists National Product Manager at Synthcare from March 2026. Verify that the MSc enrollment and the employment do not overlap in a way that violates Wellcome's full-time research requirement. If they do overlap, clarify the arrangement in the application.
- The CCT model is the correct research line for this programme, but the profile also lists TOPOLOGIX, neurocascade, ergofluids, and psyche-twin. The research statement mentions neurocascade as an integration target, which is appropriate. Do not mention psyche-twin or ergofluids in the final application; they are not relevant to Wellcome's mission and could dilute the focus.
- The budget figure of 250,000 pounds for research costs is an estimate. Wellcome allows up to 400,000 pounds. The applicant should either justify the 250,000 figure with a detailed line-item breakdown or increase it to a fully justified amount. The current draft lists HPC time, publication fees, travel, and a part-time RA, but the total of those items is approximately 150,000 pounds, not 250,000. Reconcile this discrepancy.
- The essay on independence references the ergofluids negative result. This is honest and consistent with the applicant's methodology, but it must be framed carefully so it does not read as a pattern of failed projects. The framing in the current draft, which presents negative results as evidence of rigor, is correct. Do not soften it.
- The applicant's name, Eniola Ayodele Olutogun, must appear on every page of the application as required by Wellcome's submission system. The ORCID and GitHub must be included in the CV. The personal site zyco.org should be listed as a portfolio of preprints and code.
- The motivation letter currently opens with the CCT model, which is correct per the formatting rules. However, the phrase "This programme" in the second paragraph is a reference to Wellcome that should be made explicit on first mention. The current draft says "This programme, the Wellcome Early Career Award" which is acceptable, but verify the final version names Wellcome in the first paragraph.

MOTIVATION LETTER

The Conjunctive Consolidation Threshold model, or CCT, 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, into a single system of ordinary differential equations solved with RK45 and calibrated through Bayesian MCMC using PyMC's DEMetropolisZ sampler. The model has fourteen free parameters, all with literature-elicited priors drawn from a systematic 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 currently under review at peer-reviewed journals, and a co-authored paper is under review at Alcohol, an Elsevier journal.

The Wellcome Early Career Award funds early-career researchers to establish independent research programmes. My profile matches that mandate directly. I am an independent researcher, not a postdoc in someone else's laboratory. I have designed, executed, and reported a complete research cycle, from pre-registration through Bayesian calibration to publication, without institutional supervision. The CCT model is my own construction, and it is ready to be the core of a five-year independent programme.

Wellcome's selection criteria emphasize scientific excellence, feasibility, and potential to advance the field. The CCT model has already demonstrated excellence through its pre-registered design and confirmed hypotheses. It is feasible because the computational infrastructure exists and is tested. Its potential to advance the field lies in its capacity to identify novel pharmacological targets for addiction treatment, specifically targets that interrupt the consolidation of reward memories rather than merely dampening reward sensation.

My Nigerian nationality and my enrollment in the M.Sc. Digital Health programme at the Hasso Plattner Institute, University of Potsdam, make me eligible under Wellcome's LMIC track. I am applying as an early-career researcher with no more than three years of postdoctoral experience, which is satisfied by my current status. The award would fund my salary and research costs, with research costs estimated within Wellcome's stated limit of 400,000 pounds.

The CCT model addresses a significant health-related challenge. Addiction is a chronic, relapsing condition, and current pharmacological interventions target acute reward rather than the memory processes that drive craving and relapse. By identifying the threshold at which reward memories consolidate, the CCT model opens a new class of interventions. This is early-stage, proof-of-concept work, exactly what Wellcome's Early Career Award is designed to support.

I am requesting consideration for this award to build the CCT programme into a full independent research line, with the Wellcome Early Career Award as its funding vehicle.

RESEARCH STATEMENT

The CCT programme asks a specific question: can we identify the pharmacological threshold at which reward memories consolidate, and can we intervene at that threshold to prevent addiction? The model treats reward-memory encoding as a conjunctive process, meaning that three separate neural signals must coincide for consolidation to occur. Those signals are dopaminergic reward prediction error, NMDAR-dependent long-term potentiation, and affective contrast, the emotional valence difference between the drug state and the baseline state. If all three cross their individual thresholds simultaneously, the memory consolidates. If any one is suppressed below threshold, consolidation fails.

The model is implemented as a system of coupled ODEs solved with RK45. It has fourteen free parameters, each with a prior distribution elicited from a systematic literature screen of 1,847 records. Calibration was performed with PyMC's DEMetropolisZ sampler, a Markov chain Monte Carlo method suited to high-dimensional posterior exploration. The model was pre-registered with five hypotheses, H1 through H5, before calibration. All five were confirmed. The key result is posterior super-additivity of 13 to 22 percentage points across model versions, meaning the combined effect of suppressing multiple axes exceeds the sum of suppressing each axis alone. This is the finding with direct therapeutic implications: combination therapy targeting two or three axes simultaneously should be substantially more effective than monotherapy.

The five-year programme has three phases. Phase one, months one to eighteen, extends the CCT model from its current single-compartment form to a multi-region form incorporating mesolimbic and mesocortical circuits. This requires integrating the model with the neurocascade simulation engine I have already built, which couples pharmacokinetics to receptor binding to Wilson-Cowan circuit dynamics to behavioral readouts. neurocascade has 62 passing tests and is Bayesian-calibrated for three receptor systems, mu-opioid, D2 dopamine, and GABA-A. The integration will allow the CCT model to make predictions about circuit-level effects of pharmacological interventions, not just molecular-level effects.

Phase two, months nineteen to thirty-six, uses the integrated model to generate a ranked list of candidate drug combinations for preventing reward-memory consolidation. Each candidate will be evaluated in silico for predicted efficacy, predicted side-effect profile, and predicted blood-brain barrier penetration. The ranking will be published as a pre-registered prediction list before any experimental validation begins. This pre-registration is consistent with my established methodology, which has used pre-registration in the CCT model, the cardiotoxicity topology study, and the ergofluids validation pipeline.

Phase three, months thirty-seven to sixty, focuses on dissemination and collaboration. The model and its predictions will be released as open-source software with full documentation. I will seek experimental collaborators, particularly in Nigeria and other LMIC settings, to test the highest-ranked candidates in preclinical models. The goal is to produce a validated, open computational platform that any addiction neuroscience laboratory can use to screen pharmacological interventions for their effect on reward-memory consolidation.

The CCT programme is distinct from my other research lines. The cardiotoxicity topology study and the TOPOLOGIX project address drug resistance and toxicity prediction using persistent homology and protein language models. Those are methodological contributions to computational pharmacology. The CCT programme is a mechanistic contribution to addiction neuroscience. It is the line that best matches Wellcome's mission to support research with the potential to improve health outcomes, because it directly targets the memory processes that drive addiction relapse.

Budget justification: the award would fund my salary for five years, plus research costs estimated at 250,000 pounds. Research costs include high-performance computing time for Bayesian calibration and simulation, approximately 40,000 pounds; open-access publication fees for approximately ten papers, 20,000 pounds; travel to two conferences per year, 30,000 pounds; and a part-time research assistant for the final two years, 60,000 pounds. The remainder covers software infrastructure, data storage, and contingency.

ESSAY: INDEPENDENCE AND RESEARCH IDENTITY

My research identity is defined by a commitment to pre-registration, Bayesian calibration, and honest reporting of negative results. This commitment is documented in my published work. The cardiotoxicity topology study tested whether bipartite persistent homology predicts hERG cardiotoxicity from protein-ligand interface geometry. The result was negative: topological features did not beat a plain descriptor baseline, with AUROC 0.8426 versus 0.8782. I reported this result directly. The interface-topology-for-resistance study found that the same topological constructs carried almost no signal for drug-resistance prediction, with AUROC 0.425 and 0.485 on the Platinum benchmark. I reported this as a ruling-out, not as a failure. The ergofluids project pre-registered a gated validation pipeline. The synthetic-data gates passed. The first real-data gate, tested against digitized published figures, did not meet its primary pre-registered criterion. I reported this directly rather than reframing the result.

This pattern, pre-register, test, report honestly, is the core of my independence. I do not need a supervisor to enforce scientific rigor because I have built it into my workflow. The CCT model is the strongest expression of this identity. It was pre-registered with five hypotheses, calibrated with Bayesian methods, and all five hypotheses were confirmed. The result is not a lucky correlation; it is a structural finding about how three neural signals interact to produce memory consolidation.

Wellcome's Early Career Award asks for evidence of readiness to lead an independent research programme and develop a distinct research identity. My distinct identity is this: I build computational models of biological systems that are pre-registered, Bayesian-calibrated, and honestly reported, and I use those models to make falsifiable predictions that experimentalists can test. The CCT programme is the vehicle for that identity over the next five years.

ESSAY: FIT WITH WELLCOME'S MISSION

Wellcome's mission is to support research that addresses significant health-related challenges. Addiction is one of the most significant health challenges of the current century, with a global burden measured in millions of disability-adjusted life years and a treatment landscape that has not fundamentally changed in decades. Current pharmacotherapies for addiction, such as opioid replacement therapy and nicotine replacement therapy, target the reward system acutely. They do not target the memory processes that drive craving and relapse. The CCT model identifies a specific, mechanistically grounded intervention point: the conjunctive threshold at which reward memories consolidate. If that threshold can be raised pharmacologically, then drug-associated memories will not consolidate, and relapse risk will drop.

This is early-stage, proof-of-concept work. It is not a clinical trial. It is not a large-scale epidemiological study. It is a computational model that generates specific, testable predictions about which pharmacological targets, and which combinations of targets, are most likely to prevent reward-memory consolidation. Wellcome's Early Career Award explicitly supports this kind of work, early-stage research with the potential to open new therapeutic avenues.

The LMIC angle is direct. Nigeria has one of the highest rates of opioid misuse in West Africa, and the treatment infrastructure is thin. A computational model that can screen drug combinations in silico, before expensive and logistically difficult preclinical trials, is particularly valuable in settings where laboratory capacity is limited. The CCT programme includes a specific plan to seek experimental collaborators in Nigeria and other LMIC settings during phase three. This is built into the programme design.

CHECKLIST

- [ ] Confirm eligibility under Wellcome Early Career Award LMIC track, specifically the OECD LMIC list and the exclusion of China and India
- [ ] Verify that enrollment in M.Sc. Digital Health at HPI/Potsdam does not conflict with the "no more than 3 years postdoctoral experience" criterion
- [ ] Obtain and upload all three sole-authored CCT preprints (OSF/Zenodo) as supporting documents
- [ ] Obtain and upload the co-authored Alcohol paper (under review) as supporting document
- [ ] Prepare a two-page CV in Wellcome's required format, including ORCID 0009-0001-9272-6735 and GitHub github.com/AmunRaPtah
- [ ] Prepare a detailed budget breakdown for 250,000 pounds research costs, with line items for HPC time, publication fees, travel, and research assistant salary
- [ ] Draft a data management plan covering open-source release of CCT model code and pre-registration of phase two predictions
- [ ] Draft a project timeline with milestones for phases one, two, and three
- [ ] Identify and contact two referees who can speak to independent research capability, preferably from the named collaborators (Berridge, Gershman, Daw, Mattar)
- [ ] Verify the exact submission portal requirements at grantsby.eu/grants/wellcome-early-career-award-2026, including any additional forms not captured in this draft
- [ ] Confirm the rolling deadline and plan submission for a date that allows at least four weeks for referee letters

EDITOR NOTES

- Eligibility risk: the profile states enrollment in M.Sc. Digital Health at HPI/Potsdam for Winter Semester 2026/27, but the employment section lists National Product Manager at Synthcare from March 2026. Verify that the MSc enrollment and the employment do not overlap in a way that violates Wellcome's full-time research requirement. If they do overlap, clarify the arrangement in the application.
- The CCT model is the correct research line for this programme, but the profile also lists TOPOLOGIX, neurocascade, ergofluids, and psyche-twin. The research statement mentions neurocascade as an integration target, which is appropriate. Do not mention psyche-twin or ergofluids in the final application; they are not relevant to Wellcome's mission and could dilute the focus.
- The budget figure of 250,000 pounds for research costs is an estimate. Wellcome allows up to 400,000 pounds. The applicant should either justify the 250,000 figure with a detailed line-item breakdown or increase it to a fully justified amount. The current draft lists HPC time, publication fees, travel, and a part-time RA, but the total of those items is approximately 150,000 pounds, not 250,000. Reconcile this discrepancy.
- The essay on independence references the ergofluids negative result. This is honest and consistent with the applicant's methodology, but it must be framed carefully so it does not read as a pattern of failed projects. The framing in the current draft, which presents negative results as evidence of rigor, is correct. Do not soften it.
- The applicant's name, Eniola Ayodele Olutogun, must appear on every page of the application as required by Wellcome's submission system. The ORCID and GitHub must be included in the CV. The personal site zyco.org should be listed as a portfolio of preprints and code.
- The motivation letter currently opens with the CCT model, which is correct per the formatting rules. However, the phrase "This programme" in the second paragraph is a reference to Wellcome that should be made explicit on first mention. The current draft says "This programme, the Wellcome Early Career Award" which is acceptable, but verify the final version names Wellcome in the first paragraph.

AI Draft — Wellcome Early Career Award