Draft — SDG Innovation Summit Malaysia 2026 Scholarship

MOTIVATION LETTER

The World Health Organization estimates that antimicrobial resistance will claim ten million lives per year by 2050, with sub-Saharan Africa carrying a disproportionate share of that burden. My research addresses this crisis directly. I am Eniola Ayodele Olutogun, a Nigerian pharmacist and computational researcher, and I have built TOPOLOGIX, a machine-learning system that predicts drug-resistance mutations from protein sequence alone. The system achieves an AUROC of 0.804 plus or minus 0.025 on the Platinum benchmark of 553 mutations, outperforming structure-based tools like mCSM-lig at approximately 0.70 while covering 100 percent of mutations compared to roughly 18 percent for structure-limited methods. This means clinicians and researchers can identify resistance risks even when no protein structure exists, which is the common case for newly emerging pathogens.

The SDG Innovation Summit Malaysia 2026 selects young leaders who demonstrate commitment to sustainable development through concrete innovation. My work maps directly onto SDG 3, Good Health and Well-Being, by attacking the root cause of treatment failure in infectious disease, and SDG 9, Industry, Innovation, and Infrastructure, by demonstrating that open-source AI methods can outperform expensive structural biology pipelines. I am currently enrolled in the M.Sc. Digital Health program at the Hasso Plattner Institute and University of Potsdam, and I have published three sole-authored preprints on computational pharmacology, each under peer review at international journals. My independent research record includes a pre-registered replication study on cardiotoxicity prediction that settled a question the literature had never actually tested, and a Bayesian-calibrated model of reward-memory encoding in addiction with all five pre-registered hypotheses confirmed.

What the summit offers that my laboratory work cannot is the translation layer. I have built the models, validated them against public benchmarks, and published the negative results that disciplined my approach. What I lack is the network of policy makers, public-health practitioners, and fellow innovators across Asia and Africa who can help me understand how resistance prediction tools actually reach the clinics that need them. The summit's leadership training and cross-cultural collaboration structure would give me that exposure. I come from Nigeria, where antimicrobial resistance surveillance infrastructure is thin and where a sequence-based prediction tool could be deployed at low cost. I want to leave the summit with a concrete plan for piloting TOPOLOGIX in a real clinical or public-health setting, informed by peers working on similar deployment challenges in Southeast Asia.

I am applying for the opportunity to test my assumptions against a global cohort of SDG-focused researchers and to return to my work with a sharper sense of how computational pharmacology can serve health systems rather than just journals.

RESEARCH STATEMENT

TOPOLOGIX is a computational system that predicts drug-resistance mutations from protein sequence alone, using protein language model embeddings combined with molecular fingerprints. The system addresses a specific failure in current resistance prediction: structure-based tools require crystallized protein-ligand complexes, which exist for only a small fraction of clinically relevant mutations. When a new resistance mutation appears in a pathogen, clinicians often have sequence data within days but structural data within months or years. TOPOLOGIX closes that gap.

The technical architecture is straightforward. I use ESM-2 protein language model delta-embeddings to represent the mutation's effect on the protein sequence, Morgan/ECFP circular fingerprints to represent the drug, and a Random Forest classifier to predict whether the mutation confers resistance. On the Platinum benchmark of 553 mutations, the system achieves an AUROC of 0.804 with a standard deviation of 0.025. On SKEMPI 2.0, a binding-affinity benchmark, it achieves 0.634. These numbers matter because they beat structure-based baselines while requiring no structural input at all. The coverage advantage is decisive: TOPOLOGIX scores 100 percent of mutations, while mCSM-lig, a widely used structure-based tool, covers only about 18 percent.

The research path to this result included a critical negative finding. Before building TOPOLOGIX, I tested whether bipartite persistent homology, a topological data analysis method, could predict drug resistance from protein-ligand interface geometry. The pre-registered, powered replication on the Platinum benchmark produced AUROC values of 0.425 and 0.485, no better than chance. This negative result ruled out interface geometry as the driver of resistance and motivated the shift to sequence-representation methods. I published the finding directly rather than reframing it, because knowing what does not work is as valuable as knowing what does.

The current system has limitations that I state plainly. The Random Forest classifier is interpretable but may not capture complex epistatic interactions between multiple mutations. The SKEMPI 2.0 performance of 0.634 suggests that transfer from resistance classification to binding-affinity prediction is not automatic. The training data are drawn from public benchmarks that may not reflect the mutation distributions circulating in real clinical settings, particularly in West Africa where I am based. These limitations define the next phase of work: collecting or curating resistance mutation data from African clinical isolates, fine-tuning the ESM-2 embeddings on domain-specific sequences, and validating the model against prospective rather than retrospective benchmarks.

The SDG relevance is direct. Antimicrobial resistance is a global health threat that disproportionately affects low- and middle-income countries, where diagnostic infrastructure is weakest and where empirical prescribing is most common. A tool that predicts resistance from sequence data alone, without requiring structural biology capacity, is a tool that can be deployed in precisely those settings. The methodology is open, reproducible, and built on publicly available components. I have published the pre-registration and the negative results that shaped the design, and I will publish the full pipeline and evaluation code alongside the final results.

My broader research program spans computational pharmacology, from a Bayesian-calibrated model of reward-memory encoding in addiction to a receptor-to-behavior brain-circuit simulation engine. TOPOLOGIX is the line of work I am bringing to this summit because it is the most mature, the most directly tied to global health outcomes, and the most accessible to a non-specialist audience. It is also the line of work where an African perspective on deployment is not an afterthought but a core design constraint.

ESSAY: LEADERSHIP AND SDG COMMITMENT

My leadership is demonstrated through independent research execution rather than organizational title. Since 2024, I have designed, pre-registered, executed, and published three computational research projects as a sole investigator, including a Bayesian MCMC calibration with 14 free parameters and a literature-elicited prior screen of 1,847 records. That work required managing my own timeline, securing computational resources on HPC clusters, and navigating peer review at international journals without institutional backing. When my cardiotoxicity replication study produced a negative result, I reported it directly rather than reframing it, because scientific integrity is the foundation of any contribution to sustainable development.

My commitment to SDG 3 is rooted in my clinical training. I am a licensed pharmacist with a B.Pharm from the University of Ibadan, and I have worked as a clinical pharmacist in Nigeria. I have seen what happens when a patient's infection does not respond to first-line treatment and the laboratory cannot tell us why. TOPOLOGIX is my attempt to build the tool I wished I had in that clinic. The summit's emphasis on young leaders and changemakers matches my situation: I am 29 years old, I have a working system with published benchmark results, and I am at the stage where the right network and training could help me translate that system into a deployment pilot rather than another paper.

ESSAY: DIVERSITY AND CROSS-CULTURAL COLLABORATION

I bring a perspective that is underrepresented in computational biology: a Nigerian pharmacist working on protein language models and machine learning. My research training happened across three continents, through the University of Ibadan, the Hasso Plattner Institute in Germany, and collaborations with researchers at Michigan, Harvard, Princeton, and NYU. That mix has taught me that the same scientific question looks different depending on where you stand. A resistance prediction tool is an academic exercise in a lab with structural biology infrastructure. In a Nigerian clinic, it is a decision-support system for a doctor who cannot wait six months for a crystal structure.

The summit's selection criteria emphasize geographic and disciplinary diversity. I offer both. I also offer a track record of honest collaboration: I have shared pre-registrations publicly, published negative results, and built my systems on open-source components so that others can reproduce and extend them. I am applying to the SDG Innovation Summit Malaysia 2026 because it brings together researchers, entrepreneurs, and policy makers from regions that face similar health-system constraints. I want to learn from peers who have deployed digital health tools in Southeast Asia, and I want to contribute my experience building low-infrastructure computational tools to their projects in return.

CHECKLIST

- [ ] Confirm SDG Innovation Summit Malaysia 2026 application deadline and submission portal at https://thegyn.org/sism-fully-funded-scholarship/
- [ ] Verify eligibility for Nigerian nationals and independent researchers not affiliated with a university or NGO
- [ ] Confirm whether the scholarship covers travel, accommodation, and registration fees in full
- [ ] Prepare CV in the format required by the application portal, including ORCID 0009-0001-9272-6735 and GitHub github.com/AmunRaPtah
- [ ] Prepare academic transcripts for B.Pharm from University of Ibadan and enrollment documentation for M.Sc. Digital Health at HPI/Potsdam
- [ ] Prepare two reference letters, one from a research collaborator (Kent Berridge, Samuel Gershman, Nathaniel Daw, or Marcelo Mattar) and one from a professional supervisor
- [ ] Prepare a one-page summary of TOPOLOGIX with benchmark results, suitable for a non-specialist audience
- [ ] Prepare a one-page summary of the negative result on interface-topology-for-resistance, demonstrating scientific rigor
- [ ] Prepare a short biography (100-150 words) for the application portal
- [ ] Prepare a photograph in the format specified by the application portal
- [ ] Verify whether the application requires a letter of recommendation from a professor or employer
- [ ] Verify whether the application requires proof of English proficiency or any language certification
- [ ] Confirm whether the application requires a pitch deck or video presentation in addition to written materials
- [ ] Submit application before the deadline and save confirmation email

EDITOR NOTES

- Eligibility risk: The programme is hosted by The Global Youth Network (thegyn.org), and the applicant must confirm that independent researchers are eligible. Some youth summit scholarships require affiliation with a university, NGO, or registered organization. The applicant is enrolled in M.Sc. Digital Health at HPI/Potsdam starting Winter Semester 2026/27, which may satisfy affiliation requirements, but this must be verified against the programme terms.
- Fact verification needed: The applicant's employment timeline lists National Product Manager at Synthcare from March 2026, which is in the future relative to the profile's stated age of 29. The dates in the employment section should be checked for accuracy before submission.
- Gap to fill: The applicant must insert personal details about why Malaysia specifically, and what they hope to gain from the Southeast Asian context. The current draft references Southeast Asia generally but does not name any Malaysian institution, organization, or contact. If the applicant has any existing connection to Malaysia or Southeast Asia, it should be added.
- Gap to fill: The applicant should prepare a specific example of how TOPOLOGIX could be piloted in a real clinical or public-health setting, including a named partner or institution if possible. The current draft mentions this aspiration but does not name a concrete pilot site.
- Tone check: The motivation letter opens with a WHO statistic, which is appropriate for a non-specialist audience. The research statement is more technical and should be adjusted if the application portal indicates a different audience for that section.

MOTIVATION LETTER

The World Health Organization estimates that antimicrobial resistance will claim ten million lives per year by 2050, with sub-Saharan Africa carrying a disproportionate share of that burden. My research addresses this crisis directly. I am Eniola Ayodele Olutogun, a Nigerian pharmacist and computational researcher, and I have built TOPOLOGIX, a machine-learning system that predicts drug-resistance mutations from protein sequence alone. The system achieves an AUROC of 0.804 plus or minus 0.025 on the Platinum benchmark of 553 mutations, outperforming structure-based tools like mCSM-lig at approximately 0.70 while covering 100 percent of mutations compared to roughly 18 percent for structure-limited methods. This means clinicians and researchers can identify resistance risks even when no protein structure exists, which is the common case for newly emerging pathogens.

The SDG Innovation Summit Malaysia 2026 selects young leaders who demonstrate commitment to sustainable development through concrete innovation. My work maps directly onto SDG 3, Good Health and Well-Being, by attacking the root cause of treatment failure in infectious disease, and SDG 9, Industry, Innovation, and Infrastructure, by demonstrating that open-source AI methods can outperform expensive structural biology pipelines. I am currently enrolled in the M.Sc. Digital Health program at the Hasso Plattner Institute and University of Potsdam, and I have published three sole-authored preprints on computational pharmacology, each under peer review at international journals. My independent research record includes a pre-registered replication study on cardiotoxicity prediction that settled a question the literature had never actually tested, and a Bayesian-calibrated model of reward-memory encoding in addiction with all five pre-registered hypotheses confirmed.

What the summit offers that my laboratory work cannot is the translation layer. I have built the models, validated them against public benchmarks, and published the negative results that disciplined my approach. What I lack is the network of policy makers, public-health practitioners, and fellow innovators across Asia and Africa who can help me understand how resistance prediction tools actually reach the clinics that need them. The summit's leadership training and cross-cultural collaboration structure would give me that exposure. I come from Nigeria, where antimicrobial resistance surveillance infrastructure is thin and where a sequence-based prediction tool could be deployed at low cost. I want to leave the summit with a concrete plan for piloting TOPOLOGIX in a real clinical or public-health setting, informed by peers working on similar deployment challenges in Southeast Asia.

I am applying for the opportunity to test my assumptions against a global cohort of SDG-focused researchers and to return to my work with a sharper sense of how computational pharmacology can serve health systems rather than just journals.

RESEARCH STATEMENT

TOPOLOGIX is a computational system that predicts drug-resistance mutations from protein sequence alone, using protein language model embeddings combined with molecular fingerprints. The system addresses a specific failure in current resistance prediction: structure-based tools require crystallized protein-ligand complexes, which exist for only a small fraction of clinically relevant mutations. When a new resistance mutation appears in a pathogen, clinicians often have sequence data within days but structural data within months or years. TOPOLOGIX closes that gap.

The technical architecture is straightforward. I use ESM-2 protein language model delta-embeddings to represent the mutation's effect on the protein sequence, Morgan/ECFP circular fingerprints to represent the drug, and a Random Forest classifier to predict whether the mutation confers resistance. On the Platinum benchmark of 553 mutations, the system achieves an AUROC of 0.804 with a standard deviation of 0.025. On SKEMPI 2.0, a binding-affinity benchmark, it achieves 0.634. These numbers matter because they beat structure-based baselines while requiring no structural input at all. The coverage advantage is decisive: TOPOLOGIX scores 100 percent of mutations, while mCSM-lig, a widely used structure-based tool, covers only about 18 percent.

The research path to this result included a critical negative finding. Before building TOPOLOGIX, I tested whether bipartite persistent homology, a topological data analysis method, could predict drug resistance from protein-ligand interface geometry. The pre-registered, powered replication on the Platinum benchmark produced AUROC values of 0.425 and 0.485, no better than chance. This negative result ruled out interface geometry as the driver of resistance and motivated the shift to sequence-representation methods. I published the finding directly rather than reframing it, because knowing what does not work is as valuable as knowing what does.

The current system has limitations that I state plainly. The Random Forest classifier is interpretable but may not capture complex epistatic interactions between multiple mutations. The SKEMPI 2.0 performance of 0.634 suggests that transfer from resistance classification to binding-affinity prediction is not automatic. The training data are drawn from public benchmarks that may not reflect the mutation distributions circulating in real clinical settings, particularly in West Africa where I am based. These limitations define the next phase of work: collecting or curating resistance mutation data from African clinical isolates, fine-tuning the ESM-2 embeddings on domain-specific sequences, and validating the model against prospective rather than retrospective benchmarks.

The SDG relevance is direct. Antimicrobial resistance is a global health threat that disproportionately affects low- and middle-income countries, where diagnostic infrastructure is weakest and where empirical prescribing is most common. A tool that predicts resistance from sequence data alone, without requiring structural biology capacity, is a tool that can be deployed in precisely those settings. The methodology is open, reproducible, and built on publicly available components. I have published the pre-registration and the negative results that shaped the design, and I will publish the full pipeline and evaluation code alongside the final results.

My broader research program spans computational pharmacology, from a Bayesian-calibrated model of reward-memory encoding in addiction to a receptor-to-behavior brain-circuit simulation engine. TOPOLOGIX is the line of work I am bringing to this summit because it is the most mature, the most directly tied to global health outcomes, and the most accessible to a non-specialist audience. It is also the line of work where an African perspective on deployment is not an afterthought but a core design constraint.

ESSAY: LEADERSHIP AND SDG COMMITMENT

My leadership is demonstrated through independent research execution rather than organizational title. Since 2024, I have designed, pre-registered, executed, and published three computational research projects as a sole investigator, including a Bayesian MCMC calibration with 14 free parameters and a literature-elicited prior screen of 1,847 records. That work required managing my own timeline, securing computational resources on HPC clusters, and navigating peer review at international journals without institutional backing. When my cardiotoxicity replication study produced a negative result, I reported it directly rather than reframing it, because scientific integrity is the foundation of any contribution to sustainable development.

My commitment to SDG 3 is rooted in my clinical training. I am a licensed pharmacist with a B.Pharm from the University of Ibadan, and I have worked as a clinical pharmacist in Nigeria. I have seen what happens when a patient's infection does not respond to first-line treatment and the laboratory cannot tell us why. TOPOLOGIX is my attempt to build the tool I wished I had in that clinic. The summit's emphasis on young leaders and changemakers matches my situation: I am 29 years old, I have a working system with published benchmark results, and I am at the stage where the right network and training could help me translate that system into a deployment pilot rather than another paper.

ESSAY: DIVERSITY AND CROSS-CULTURAL COLLABORATION

I bring a perspective that is underrepresented in computational biology: a Nigerian pharmacist working on protein language models and machine learning. My research training happened across three continents, through the University of Ibadan, the Hasso Plattner Institute in Germany, and collaborations with researchers at Michigan, Harvard, Princeton, and NYU. That mix has taught me that the same scientific question looks different depending on where you stand. A resistance prediction tool is an academic exercise in a lab with structural biology infrastructure. In a Nigerian clinic, it is a decision-support system for a doctor who cannot wait six months for a crystal structure.

The summit's selection criteria emphasize geographic and disciplinary diversity. I offer both. I also offer a track record of honest collaboration: I have shared pre-registrations publicly, published negative results, and built my systems on open-source components so that others can reproduce and extend them. I am applying to the SDG Innovation Summit Malaysia 2026 because it brings together researchers, entrepreneurs, and policy makers from regions that face similar health-system constraints. I want to learn from peers who have deployed digital health tools in Southeast Asia, and I want to contribute my experience building low-infrastructure computational tools to their projects in return.

CHECKLIST

- [ ] Confirm SDG Innovation Summit Malaysia 2026 application deadline and submission portal at https://thegyn.org/sism-fully-funded-scholarship/
- [ ] Verify eligibility for Nigerian nationals and independent researchers not affiliated with a university or NGO
- [ ] Confirm whether the scholarship covers travel, accommodation, and registration fees in full
- [ ] Prepare CV in the format required by the application portal, including ORCID 0009-0001-9272-6735 and GitHub github.com/AmunRaPtah
- [ ] Prepare academic transcripts for B.Pharm from University of Ibadan and enrollment documentation for M.Sc. Digital Health at HPI/Potsdam
- [ ] Prepare two reference letters, one from a research collaborator (Kent Berridge, Samuel Gershman, Nathaniel Daw, or Marcelo Mattar) and one from a professional supervisor
- [ ] Prepare a one-page summary of TOPOLOGIX with benchmark results, suitable for a non-specialist audience
- [ ] Prepare a one-page summary of the negative result on interface-topology-for-resistance, demonstrating scientific rigor
- [ ] Prepare a short biography (100-150 words) for the application portal
- [ ] Prepare a photograph in the format specified by the application portal
- [ ] Verify whether the application requires a letter of recommendation from a professor or employer
- [ ] Verify whether the application requires proof of English proficiency or any language certification
- [ ] Confirm whether the application requires a pitch deck or video presentation in addition to written materials
- [ ] Submit application before the deadline and save confirmation email

EDITOR NOTES

- Eligibility risk: The programme is hosted by The Global Youth Network (thegyn.org), and the applicant must confirm that independent researchers are eligible. Some youth summit scholarships require affiliation with a university, NGO, or registered organization. The applicant is enrolled in M.Sc. Digital Health at HPI/Potsdam starting Winter Semester 2026/27, which may satisfy affiliation requirements, but this must be verified against the programme terms.
- Fact verification needed: The applicant's employment timeline lists National Product Manager at Synthcare from March 2026, which is in the future relative to the profile's stated age of 29. The dates in the employment section should be checked for accuracy before submission.
- Gap to fill: The applicant must insert personal details about why Malaysia specifically, and what they hope to gain from the Southeast Asian context. The current draft references Southeast Asia generally but does not name any Malaysian institution, organization, or contact. If the applicant has any existing connection to Malaysia or Southeast Asia, it should be added.
- Gap to fill: The applicant should prepare a specific example of how TOPOLOGIX could be piloted in a real clinical or public-health setting, including a named partner or institution if possible. The current draft mentions this aspiration but does not name a concrete pilot site.
- Tone check: The motivation letter opens with a WHO statistic, which is appropriate for a non-specialist audience. The research statement is more technical and should be adjusted if the application portal indicates a different audience for that section.

AI Draft — SDG Innovation Summit Malaysia 2026 Scholarship