Democratizing Health Advances: Artificial Intelligence, Genomics, and Our Paths to Equity

Geoffrey Siwo recently joined the Center for Global Equity as a research associate and is helping to develop the Center’s new Data Collaborative. Siwo will focus on harmonizing, sharing, and interpreting data in new ways and ensuring data can be used by those who need it most.

We asked Dr. Siwo to share more about his academic background, especially as it relates to health equity around the world.

How do you approach global health equity—as a data scientist but also beyond your role as a researcher?

Data science can play an important role in global health equity if leveraged responsibly. An important path to health equity is democratizing who and where medical advances are made. The who is crucial. We need more diversity in the teams addressing global health challenges, just as we need diversity in everything else.

Data science can help address inequities, but it can also reproduce existing inequities or create new ones.

As a data scientist, I will say that we put too much faith in data itself without understanding the context of what is behind the data. We can’t effectively address inequities if certain populations within the full human population are underrepresented among data scientists.

Data science can help address inequities, but it can also reproduce existing inequities or create new ones. New business models are needed to develop and enhance equitable medical solutions. Translating research into real-world applications is my long-term goal. This requires entrepreneurship and partnerships that go beyond academia.

Would you tell us about a research project that illuminates the impact of your work?

Crowdsourcing can help us accelerate the development of new computational models that address global health problems. At Notre Dame, I led the Malaria DREAM challenge, which develops computational models for understanding emerging resistance to the anti-malarial drug artemisinin. Malaria drug resistance is a big global health problem that threatens to reverse decades of progress made in the fight against the disease.

The Malaria DREAM challenge was inspired by a simple idea that goes against how research typically occurs in academia and industry.

The quality and diversity of the models developed by participants surpassed what could be achieved in a single institution or company.

Conventionally, researchers in individual labs analyze their own datasets to derive new scientific insights. Or at best they collaborate with other pre-selected labs. In the Malaria DREAM challenge, we opened unpublished genome datasets from several malaria parasites to a large group of people who were not our collaborators. Then we engaged the public in two ways.

First, we organized a week-long hackathon at IBM Research Africa’s lab in Johannesburg. Twenty young African scientists from 8 countries collaboratively analyzed the data to determine if it could help us better understand artemisinin drug resistance. Unlike a collaboration between the West and Africa, where data tends to be “exported” out of Africa for analysis, the malaria hackathon data came from collaborators in South East Asia, was processed in US-based labs, and was then “exported” to Africa for analysis.

Second, we opened the data to the public and challenged anyone interested to help us explore whether the resistance level of malaria parasites can be predicted using these datasets. Over a period of 3 months, 360 individuals from 31 countries participated in the challenge.

Equity is a process and a practice more than a goal.

We placed no restrictions on participation. And the quality and diversity of the models developed by participants surpassed what could be achieved in a single institution or company. Participants ranged in expertise and experience from high school to start-up founders to university professors.

How do you define success in global health equity work?

To me success is not a destination but a path we take toward global health equity. The target of what we call “global health” is continuously evolving. So equity is about more than an equal distribution of existing medical interventions. Equity also demands a commitment to innovations that maximize benefits for all now and in the future. Equity is a process and a practice more than a goal.

This is why my research focuses on emerging, exponentially growing technologies like AI and genomics. As we see with COVID-19 vaccines and especially RNA vaccines, new technologies when successful are more likely to be applied inequitably, in part because of their high initial cost and barriers to market entry such as IP held by pioneers. Moving research toward open-source approaches to problems like vaccine access can help humanity more equitably distribute medical interventions and other health measures.

Photo above. Siwo at the official launch of the IBM Research Africa lab (Johannesburg, South Africa) with Ginni Rometty, then IBM Chair and CEO