The Genomic Architect: Bridging Field Biodiversity and Computational Precision

A conversation with Adeola Oluwakemi Ayoola, PhD — Project Scientist at UCLA's David Geffen School of Medicine — on population genomics, infectious disease surveillance, and how computational biology is reshaping our understanding of disease, evolution, and the One Health framework.

By Louise Servoin · 2026-06-09 · 8 min read

## Adeola Oluwakemi Ayoola, PhD – Project Scientist, Goldberg Lab, UCLA (Human Genetics)

Expertise: Computational Biology, Population Genomics, and Next-Generation Sequencing (NGS) Pipeline Engineering. Visionary leader bridging the gap between field-based biospecimen collection in West Africa and large-scale genomic inference using AI and HPC. Based in North Carolina.

Some scientists are defined by their ability to see the world through a single lens; others, like Dr. Adeola Ayoola, build bridges across the arc of science. Her journey began in wildlife ecology in Nigeria, before expanding that same lens inward, from ecosystems to genomes.

From her doctoral studies at the Chinese Academy of Sciences to her research at Duke University and now UCLA, she has pioneered a “bench-to-bioinformatics” approach that is reshaping how we understand zoonotic spillover, primate evolution and clinical research.

Across every stop, one question has driven her work: how do we build the genomic infrastructure, the pipelines, the protocols and the people, that actually protects health, human, animal, and environmental, at a global scale?

We met with Adeola to discuss the transition from the field to the computing cluster, the critical role of African genomics in global health, and why “cross-layer fluency” is the most vital skill for the next generation of biologists.

## The Evolution of a Perspective

Adeola, your background is fascinating. You started in Wildlife Management and Ecology in Nigeria and now lead genomic analyses at UCLA. How does that early field experience inform your work in human genetics today?

It is fundamentally about understanding the context behind the data. In my early training at the Federal University of Technology Akure, I was immersed in the macroscopic reality of biodiversity and conservation. When I moved deeper into genetics during my PhD in China, I didn't leave that behind, I simply shifted the resolution. Today, whether I am building NGS pipelines, modeling polygenic risk scores, or studying genetic adaptation at UCLA, I understand what those samples cost, the logistics, the ethics, the communities, and the conditions, across multi-site field studies in underserved communities in West Africa, working with HRH-CERID LAUTECH (https://www.linkedin.com/in/hrh-cerid-lautech-ogbomoso-5887b0240/), or alongside rangers at national parks. That grounding never left. That cross-layer fluency is what matters. It tells you why a dataset looks the way it does, what the missing data actually means, and how to translate a computational finding into something that is biologically and clinically actionable.

"Few computational genomics scientists have spent equal time at the bench, in the field, and inside HPC pipelines. That fluency is the bridge between a biological sample and a population-scale inference." — Adeola Ayoola, PhD

## The Digital Pulse of Evolution

You work daily with R, Python, and Snakemake to build reproducible pipelines. How is high-performance computing (HPC) and AI changing our ability to track infectious diseases?

We are moving from reactive monitoring to predictive modeling. At UCLA and previously at Duke, we have deployed end-to-end NGS workflows to investigate how pathogens like SIV and Plasmodium have left signatures on primate and human genomes. Using selection scans, iHS, Tajima's D, and FST, we can identify adaptive signatures, molecular scars, if you will, that tell us how populations have survived past pathogen pressure. AI and machine learning now allow us to integrate multi-omics layers, transcriptomic, epigenomic, proteomic, to evaluate how genomic variation relates to disease susceptibility and function. We are not just reading the past; we are building GWAS-ready datasets and polygenic risk frameworks that help predict future disease outcomes.

## One Health and Global Impact

A significant part of your work involves “One Health.” You’ve led projects ranging from searching for the rarest primates in Côte d'Ivoire to managing infectious disease labs in Nigeria. Why is West Africa so central to your scientific mission?

West Africa is a frontline for both biodiversity and emerging infectious diseases. My work with the Centre Suisse de Recherches Scientifiques involved using carrion flies as biological samplers to find the critically endangered Miss Waldron's Red Colobus, a non-invasive way to “listen” to the environment through DNA. But One Health is not only about surveillance. It is about building the capacity to act on what you find. When I served as Clinical Research Associate and Laboratory Manager at the Humboldt Research Hub in Nigeria in 2022, I coordinated multi-site cohort studies on Covid, Rotavirus, HIV, and malaria across underserved communities. I authored IRB-aligned SOPs, trained over 50 research personnel across multiple sites, and scaled operational readiness from 30% to full audit-ready capacity within six months. That experience taught me something that no amount of pipeline optimization can: you cannot have global health security without robust, locally-led, standardized research infrastructure in Africa. The genomic insights mean nothing if the systems to generate and act on them do not exist.

## The Golden Thread

We often see a gap between what technology can do and what professionals actually adopt. How do we ensure AI becomes a ‘co-pilot’ that elevates human judgment rather than a tool that encourages intellectual passivity?

This is the most important question in the room right now, and I speak from direct experience on both sides of it. I have used LLM-assisted coding to accelerate pipeline development and workflow automation, and I have also watched well-intentioned AI tools fail in the field because the people expected to use them were never part of their design.

The answer is not better algorithms. It is better integration. An AI tool deployed in a clinical research site in rural Nigeria, or a genomic analysis platform adopted by a hospital system in Accra, has to be designed with those users, not just for them. That means community co-design, contextual training, and building local champions who understand both the science and the human workflows it must sit inside.

At the same time, I think we have to be honest that intellectual passivity is a real risk, not because AI makes people lazy, but because it lowers the perceived cost of skipping the hard reasoning. The antidote is rigorous scientific training that teaches people to interrogate outputs, not just consume them. I mentor early-career scientists specifically to understand the biological and clinical question before they touch a model. The model should serve the question; the question should never be invented to fit the model.

In research and healthcare, the stakes of getting that backwards are not abstract. They are measured in misdiagnoses, in missed outbreak signals, in communities whose health data is used to build tools that never return to benefit them. AI as a co-pilot only works if the human in the seat actually knows how to fly.

"The model should serve the question; the question should never be invented to fit the model." — Adeola Ayoola, PhD

Tags: Population Genomics, One Health, Bioinformatics, Infectious Disease, UCLA, Computational Biology, African Genomics