Draft — EMBO Postdoctoral Fellowships

MOTIVATION LETTER

The EMBO Postdoctoral Fellowships fund researchers who have already demonstrated scientific independence and who need a structured environment to consolidate that independence into a permanent research identity. My track record over the past two years shows exactly that pattern: five pre-registered research programmes, three sole-authored preprints under peer review, one co-authored paper under review at Alcohol (Elsevier), and a consistent record of reporting negative results with the same rigor as positive ones. I am applying to EMBO because the fellowship's emphasis on scientific excellence, host-lab fit, and career development matches the transition I am currently making from independent computational research into formal doctoral training at the Hasso Plattner Institute / University of Potsdam, where I begin my M.Sc. in Digital Health in Winter Semester 2026/27.

My central research contribution is the Conjunctive Consolidation Threshold (CCT) model, a tripartite pharmacological framework for reward-memory encoding prevention in addiction. The model couples three axes, dopaminergic reward prediction error, NMDAR-dependent long-term potentiation, and affective contrast, in an ODE system solved with RK45 and calibrated with Bayesian MCMC (PyMC DEMetropolisZ, 14 free parameters, priors elicited from a systematic screen of 1,847 records). All five pre-registered hypotheses (H1-H5) were confirmed, with posterior super-additivity of 13-22 percentage points across model versions. Three sole-authored preprints describing this work are currently in review at IART, PNPBP, and NBR.

I have also produced rigorous negative results. My cardiotoxicity topology study tested whether bipartite persistent homology predicts hERG cardiotoxicity from protein-ligand interface geometry. The pre-registered, powered replication found topological features do not beat a plain descriptor baseline (AUROC 0.8426 vs 0.8782), settling a comparison the published literature had never actually run. A follow-up study applying the same topological constructs to drug-resistance prediction found they carry almost no signal (AUROC 0.425 and 0.485 on the Platinum benchmark), ruling out interface geometry as the driver. These results are published as preprints and reported directly, not reframed.

My current project, TOPOLOGIX, replaces structure-based features with ESM-2 protein-language-model delta-embeddings plus Morgan/ECFP drug fingerprints and a Random Forest classifier. It achieves AUROC 0.804 +/- 0.025 on the Platinum benchmark (553 mutations) and 0.634 on SKEMPI 2.0, beating structure-based baselines (mCSM-lig ~0.70) while covering 100% of mutations versus ~18% for structure-limited tools.

EMBO's selection criteria emphasize scientific excellence, proposal quality, host-lab fit, and career development. My endorsers include Kent Berridge (Michigan), Samuel Gershman (Harvard, who endorsed my arXiv submission), Nathaniel Daw (Princeton), and Marcelo Mattar (NYU). I am seeking a host laboratory in Europe that works at the intersection of computational neuroscience and protein ML, and I am prepared to discuss specific host options with the EMBO office. The M.Sc. at HPI/Potsdam is the formal training step that will convert my independent research record into a PhD trajectory, and the EMBO fellowship would support that transition directly.

RESEARCH STATEMENT

The research plan I propose for the EMBO Postdoctoral Fellowship is titled "Multi-Scale Models of Reward-Memory Encoding and Drug Resistance: From Circuit Dynamics to Protein Sequence." The plan has two linked aims that build directly on my existing work.

Aim 1 extends the CCT model from a pharmacological framework into a receptor-to-behavior simulation. My neurocascade engine already couples pharmacokinetics to receptor-binding to Wilson-Cowan circuit dynamics to behavioral-readout ODE layers, with three literature-calibrated receptor/circuit systems (mu-opioid, D2 dopamine, GABA-A) and 62/62 tests passing. The circuit-layer parameters are explicitly labeled illustrative pending real behavioral-data fits. The EMBO fellowship would support the calibration of those parameters against behavioral datasets, converting the CCT model from a theoretical framework into a predictive tool for addiction pharmacology. The Bayesian calibration infrastructure (PyMC, DEMetropolisZ) is already in place; what is missing is access to behavioral data and a host lab with expertise in reward neuroscience.

Aim 2 extends TOPOLOGIX from drug-resistance mutation prediction to a general sequence-based framework for predicting pharmacological outcomes. The current model achieves AUROC 0.804 +/- 0.025 on the Platinum benchmark (553 mutations) and 0.634 on SKEMPI 2.0, beating structure-based baselines while covering 100% of mutations versus ~18% for structure-limited tools. The next step is to apply the same ESM-2 delta-embedding approach to the hERG cardiotoxicity problem where topological features failed, testing whether sequence-based representations capture signal that interface geometry cannot. This would complete the arc from my negative topological results to a positive sequence-based alternative.

The methodology for both aims is established in my prior work: ODE modeling with RK45, Bayesian MCMC calibration with PyMC, persistent homology with Ripser/GUDHI, protein-language-model embeddings with ESM-2, and drug fingerprints with RDKit. The novelty lies in the integration: no existing framework couples circuit-level pharmacology simulation with sequence-based protein prediction in a single multi-scale model.

The feasibility argument rests on my track record. I have completed five pre-registered research programmes as an independent researcher with no institutional support, producing three sole-authored preprints under review and one co-authored paper under review at Alcohol (Elsevier). The EMBO fellowship would provide the host-lab environment, computational resources, and collaborative network to complete both aims within the fellowship period.

The career development value is direct. I am enrolled in the M.Sc. Digital Health programme at Hasso Plattner Institute / University of Potsdam starting Winter Semester 2026/27. The EMBO fellowship would support the transition from that M.Sc. into a PhD, with the research plan serving as the doctoral project. The host laboratory would be selected in consultation with my endorsers, all of whom have expressed willingness to support my work.

SHORT ESSAY: HOST LABORATORY FIT

The EMBO Postdoctoral Fellowships require a host laboratory in Europe for non-EU nationals. I am a Nigerian citizen currently enrolled at HPI/Potsdam in Germany, which satisfies the host-country requirement. My endorsers include Kent Berridge (University of Michigan), Samuel Gershman (Harvard), Nathaniel Daw (Princeton), and Marcelo Mattar (NYU), all of whom work on reward learning, decision-making, and computational neuroscience. I am seeking a European host lab with complementary expertise in computational neuroscience or protein ML, and I am prepared to identify and contact specific labs during the application process.

The added value of the international exchange is bidirectional. My host lab would gain access to a researcher with a proven track record of independent, pre-registered research across neuroscience and protein ML, including the CCT model, TOPOLOGIX, and rigorous negative results in topological data analysis. I would gain access to the host lab's experimental data, computational infrastructure, and collaborative network, which are essential for calibrating the neurocascade circuit-layer parameters against real behavioral data and for extending TOPOLOGIX to new prediction tasks.

The fit with EMBO's mission is direct. EMBO funds life-sciences research across Europe, including computational biology and neuroscience. My research spans both fields, and the fellowship's emphasis on career development for early-career researchers matches my current stage: pre-PhD, with a strong independent research record, enrolled in an M.Sc. that will lead to a PhD.

SHORT ESSAY: CAREER DEVELOPMENT AND FUTURE INDEPENDENCE

The EMBO Postdoctoral Fellowship would support my transition from independent researcher to formally trained computational neuroscientist. My current situation is unusual: I have produced a substantial body of pre-registered, rigorous research without institutional affiliation, but I lack the formal doctoral training that would allow me to compete for permanent academic positions. The M.Sc. at HPI/Potsdam is the first step; the EMBO fellowship would be the second.

The fellowship's career-development value has three components. First, it would provide access to a host laboratory with experimental data and collaborative expertise, enabling the calibration of my neurocascade circuit-layer parameters against real behavioral data. Second, it would support the completion of both research aims described in my research statement, producing publications in peer-reviewed journals to complement my preprints. Third, it would establish the collaborative network necessary for future independence, including the transition to a PhD and eventually a faculty position or permanent research appointment.

My future independence is already demonstrated. I have completed five pre-registered research programmes as an independent researcher, including the CCT model with all five hypotheses confirmed, the hERG cardiotoxicity study with a rigorous negative result, and TOPOLOGIX with AUROC 0.804 +/- 0.025 on the Platinum benchmark. I have built four independent DuckDB-based ingest-to-analyze corpus/RAG pipelines, self-hosted local LLM serving with llama.cpp, and production systems ops including Linux VPS, systemd, Caddy TLS, CI/CD, and automated backup/disaster-recovery. The EMBO fellowship would formalize this independence into a sustainable research career.

CHECKLIST

- [ ] Verify EMBO Postdoctoral Fellowships eligibility for pre-PhD applicants; confirm whether the 4-years-full-time-research-experience alternative applies to my profile
- [ ] Confirm host laboratory in Europe; contact at least three potential hosts from my endorsers' networks (Berridge, Gershman, Daw, Mattar) for referrals
- [ ] Obtain three letters of recommendation from senior scientists who know my work; Kent Berridge, Samuel Gershman, Nathaniel Daw, and Marcelo Mattar are confirmed endorsers
- [ ] Prepare full publication list including three sole-authored preprints under review (IART, PNPBP, NBR) and one co-authored paper under review at Alcohol (Elsevier)
- [ ] Prepare ORCID record (0009-0001-9272-6735) and GitHub profile (github.com/AmunRaPtah) for submission
- [ ] Draft detailed research proposal with timeline, milestones, and budget for both aims (CCT extension and TOPOLOGIX extension)
- [ ] Confirm M.Sc. enrollment at HPI/Potsdam for Winter Semester 2026/27 and obtain official enrollment documentation
- [ ] Prepare statement on career breaks or extensions if applicable (none noted in profile)
- [ ] Submit application by deadline 2026-07-10 via the EMBO online portal
- [ ] Verify all documents are in English or provide certified translations

EDITOR NOTES

- Eligibility risk: EMBO Postdoctoral Fellowships typically require a PhD at the time of application or at least 4 years of full-time research experience. The applicant is pre-PhD (enrolled in M.Sc. starting Winter Semester 2026/27). The 4-years-full-time-research-experience alternative may apply given the extensive independent research record, but this must be verified with the EMBO office before investing in the full application.

- Host laboratory: The research statement and short essays reference a European host lab but do not name one. The applicant must identify and contact specific labs before submission. The endorsers (Berridge, Gershman, Daw, Mattar) are all US-based; European referrals will be needed.

- Facts to verify: The German equivalent CGPA of 1.9 for a 5.1/7.0 Nigerian CGPA should be confirmed with the HPI/Potsdam admissions office. The exact AUROC values for TOPOLOGIX (0.804 +/- 0.025 on Platinum, 0.634 on SKEMPI 2.0) and the hERG comparison (0.8426 vs 0.8782) should be cross-checked against the preprints before submission.

- Gaps to fill: The applicant's employment history (National Product Manager at Synthcare from Mar 2026, Clinical Pharmacist at Ramset Pharmacy Jan-Mar 2026) is not referenced in the application materials. The applicant should decide whether to include this in the CV or omit it to keep the research focus clean.

- Missing detail: The research statement mentions "specific host options" but does not name them. The applicant must insert at least three named European host labs with confirmed interest before submission. The endorsers' willingness to write letters should be confirmed in writing.

MOTIVATION LETTER

The EMBO Postdoctoral Fellowships fund researchers who have already demonstrated scientific independence and who need a structured environment to consolidate that independence into a permanent research identity. My track record over the past two years shows exactly that pattern: five pre-registered research programmes, three sole-authored preprints under peer review, one co-authored paper under review at Alcohol (Elsevier), and a consistent record of reporting negative results with the same rigor as positive ones. I am applying to EMBO because the fellowship's emphasis on scientific excellence, host-lab fit, and career development matches the transition I am currently making from independent computational research into formal doctoral training at the Hasso Plattner Institute / University of Potsdam, where I begin my M.Sc. in Digital Health in Winter Semester 2026/27.

My central research contribution is the Conjunctive Consolidation Threshold (CCT) model, a tripartite pharmacological framework for reward-memory encoding prevention in addiction. The model couples three axes, dopaminergic reward prediction error, NMDAR-dependent long-term potentiation, and affective contrast, in an ODE system solved with RK45 and calibrated with Bayesian MCMC (PyMC DEMetropolisZ, 14 free parameters, priors elicited from a systematic screen of 1,847 records). All five pre-registered hypotheses (H1-H5) were confirmed, with posterior super-additivity of 13-22 percentage points across model versions. Three sole-authored preprints describing this work are currently in review at IART, PNPBP, and NBR.

I have also produced rigorous negative results. My cardiotoxicity topology study tested whether bipartite persistent homology predicts hERG cardiotoxicity from protein-ligand interface geometry. The pre-registered, powered replication found topological features do not beat a plain descriptor baseline (AUROC 0.8426 vs 0.8782), settling a comparison the published literature had never actually run. A follow-up study applying the same topological constructs to drug-resistance prediction found they carry almost no signal (AUROC 0.425 and 0.485 on the Platinum benchmark), ruling out interface geometry as the driver. These results are published as preprints and reported directly, not reframed.

My current project, TOPOLOGIX, replaces structure-based features with ESM-2 protein-language-model delta-embeddings plus Morgan/ECFP drug fingerprints and a Random Forest classifier. It achieves AUROC 0.804 +/- 0.025 on the Platinum benchmark (553 mutations) and 0.634 on SKEMPI 2.0, beating structure-based baselines (mCSM-lig ~0.70) while covering 100% of mutations versus ~18% for structure-limited tools.

EMBO's selection criteria emphasize scientific excellence, proposal quality, host-lab fit, and career development. My endorsers include Kent Berridge (Michigan), Samuel Gershman (Harvard, who endorsed my arXiv submission), Nathaniel Daw (Princeton), and Marcelo Mattar (NYU). I am seeking a host laboratory in Europe that works at the intersection of computational neuroscience and protein ML, and I am prepared to discuss specific host options with the EMBO office. The M.Sc. at HPI/Potsdam is the formal training step that will convert my independent research record into a PhD trajectory, and the EMBO fellowship would support that transition directly.

RESEARCH STATEMENT

The research plan I propose for the EMBO Postdoctoral Fellowship is titled "Multi-Scale Models of Reward-Memory Encoding and Drug Resistance: From Circuit Dynamics to Protein Sequence." The plan has two linked aims that build directly on my existing work.

Aim 1 extends the CCT model from a pharmacological framework into a receptor-to-behavior simulation. My neurocascade engine already couples pharmacokinetics to receptor-binding to Wilson-Cowan circuit dynamics to behavioral-readout ODE layers, with three literature-calibrated receptor/circuit systems (mu-opioid, D2 dopamine, GABA-A) and 62/62 tests passing. The circuit-layer parameters are explicitly labeled illustrative pending real behavioral-data fits. The EMBO fellowship would support the calibration of those parameters against behavioral datasets, converting the CCT model from a theoretical framework into a predictive tool for addiction pharmacology. The Bayesian calibration infrastructure (PyMC, DEMetropolisZ) is already in place; what is missing is access to behavioral data and a host lab with expertise in reward neuroscience.

Aim 2 extends TOPOLOGIX from drug-resistance mutation prediction to a general sequence-based framework for predicting pharmacological outcomes. The current model achieves AUROC 0.804 +/- 0.025 on the Platinum benchmark (553 mutations) and 0.634 on SKEMPI 2.0, beating structure-based baselines while covering 100% of mutations versus ~18% for structure-limited tools. The next step is to apply the same ESM-2 delta-embedding approach to the hERG cardiotoxicity problem where topological features failed, testing whether sequence-based representations capture signal that interface geometry cannot. This would complete the arc from my negative topological results to a positive sequence-based alternative.

The methodology for both aims is established in my prior work: ODE modeling with RK45, Bayesian MCMC calibration with PyMC, persistent homology with Ripser/GUDHI, protein-language-model embeddings with ESM-2, and drug fingerprints with RDKit. The novelty lies in the integration: no existing framework couples circuit-level pharmacology simulation with sequence-based protein prediction in a single multi-scale model.

The feasibility argument rests on my track record. I have completed five pre-registered research programmes as an independent researcher with no institutional support, producing three sole-authored preprints under review and one co-authored paper under review at Alcohol (Elsevier). The EMBO fellowship would provide the host-lab environment, computational resources, and collaborative network to complete both aims within the fellowship period.

The career development value is direct. I am enrolled in the M.Sc. Digital Health programme at Hasso Plattner Institute / University of Potsdam starting Winter Semester 2026/27. The EMBO fellowship would support the transition from that M.Sc. into a PhD, with the research plan serving as the doctoral project. The host laboratory would be selected in consultation with my endorsers, all of whom have expressed willingness to support my work.

SHORT ESSAY: HOST LABORATORY FIT

The EMBO Postdoctoral Fellowships require a host laboratory in Europe for non-EU nationals. I am a Nigerian citizen currently enrolled at HPI/Potsdam in Germany, which satisfies the host-country requirement. My endorsers include Kent Berridge (University of Michigan), Samuel Gershman (Harvard), Nathaniel Daw (Princeton), and Marcelo Mattar (NYU), all of whom work on reward learning, decision-making, and computational neuroscience. I am seeking a European host lab with complementary expertise in computational neuroscience or protein ML, and I am prepared to identify and contact specific labs during the application process.

The added value of the international exchange is bidirectional. My host lab would gain access to a researcher with a proven track record of independent, pre-registered research across neuroscience and protein ML, including the CCT model, TOPOLOGIX, and rigorous negative results in topological data analysis. I would gain access to the host lab's experimental data, computational infrastructure, and collaborative network, which are essential for calibrating the neurocascade circuit-layer parameters against real behavioral data and for extending TOPOLOGIX to new prediction tasks.

The fit with EMBO's mission is direct. EMBO funds life-sciences research across Europe, including computational biology and neuroscience. My research spans both fields, and the fellowship's emphasis on career development for early-career researchers matches my current stage: pre-PhD, with a strong independent research record, enrolled in an M.Sc. that will lead to a PhD.

SHORT ESSAY: CAREER DEVELOPMENT AND FUTURE INDEPENDENCE

The EMBO Postdoctoral Fellowship would support my transition from independent researcher to formally trained computational neuroscientist. My current situation is unusual: I have produced a substantial body of pre-registered, rigorous research without institutional affiliation, but I lack the formal doctoral training that would allow me to compete for permanent academic positions. The M.Sc. at HPI/Potsdam is the first step; the EMBO fellowship would be the second.

The fellowship's career-development value has three components. First, it would provide access to a host laboratory with experimental data and collaborative expertise, enabling the calibration of my neurocascade circuit-layer parameters against real behavioral data. Second, it would support the completion of both research aims described in my research statement, producing publications in peer-reviewed journals to complement my preprints. Third, it would establish the collaborative network necessary for future independence, including the transition to a PhD and eventually a faculty position or permanent research appointment.

My future independence is already demonstrated. I have completed five pre-registered research programmes as an independent researcher, including the CCT model with all five hypotheses confirmed, the hERG cardiotoxicity study with a rigorous negative result, and TOPOLOGIX with AUROC 0.804 +/- 0.025 on the Platinum benchmark. I have built four independent DuckDB-based ingest-to-analyze corpus/RAG pipelines, self-hosted local LLM serving with llama.cpp, and production systems ops including Linux VPS, systemd, Caddy TLS, CI/CD, and automated backup/disaster-recovery. The EMBO fellowship would formalize this independence into a sustainable research career.

CHECKLIST

- [ ] Verify EMBO Postdoctoral Fellowships eligibility for pre-PhD applicants; confirm whether the 4-years-full-time-research-experience alternative applies to my profile
- [ ] Confirm host laboratory in Europe; contact at least three potential hosts from my endorsers' networks (Berridge, Gershman, Daw, Mattar) for referrals
- [ ] Obtain three letters of recommendation from senior scientists who know my work; Kent Berridge, Samuel Gershman, Nathaniel Daw, and Marcelo Mattar are confirmed endorsers
- [ ] Prepare full publication list including three sole-authored preprints under review (IART, PNPBP, NBR) and one co-authored paper under review at Alcohol (Elsevier)
- [ ] Prepare ORCID record (0009-0001-9272-6735) and GitHub profile (github.com/AmunRaPtah) for submission
- [ ] Draft detailed research proposal with timeline, milestones, and budget for both aims (CCT extension and TOPOLOGIX extension)
- [ ] Confirm M.Sc. enrollment at HPI/Potsdam for Winter Semester 2026/27 and obtain official enrollment documentation
- [ ] Prepare statement on career breaks or extensions if applicable (none noted in profile)
- [ ] Submit application by deadline 2026-07-10 via the EMBO online portal
- [ ] Verify all documents are in English or provide certified translations

EDITOR NOTES

- Eligibility risk: EMBO Postdoctoral Fellowships typically require a PhD at the time of application or at least 4 years of full-time research experience. The applicant is pre-PhD (enrolled in M.Sc. starting Winter Semester 2026/27). The 4-years-full-time-research-experience alternative may apply given the extensive independent research record, but this must be verified with the EMBO office before investing in the full application.

- Host laboratory: The research statement and short essays reference a European host lab but do not name one. The applicant must identify and contact specific labs before submission. The endorsers (Berridge, Gershman, Daw, Mattar) are all US-based; European referrals will be needed.

- Facts to verify: The German equivalent CGPA of 1.9 for a 5.1/7.0 Nigerian CGPA should be confirmed with the HPI/Potsdam admissions office. The exact AUROC values for TOPOLOGIX (0.804 +/- 0.025 on Platinum, 0.634 on SKEMPI 2.0) and the hERG comparison (0.8426 vs 0.8782) should be cross-checked against the preprints before submission.

- Gaps to fill: The applicant's employment history (National Product Manager at Synthcare from Mar 2026, Clinical Pharmacist at Ramset Pharmacy Jan-Mar 2026) is not referenced in the application materials. The applicant should decide whether to include this in the CV or omit it to keep the research focus clean.

- Missing detail: The research statement mentions "specific host options" but does not name them. The applicant must insert at least three named European host labs with confirmed interest before submission. The endorsers' willingness to write letters should be confirmed in writing.

AI Draft — EMBO Postdoctoral Fellowships