Toward Disease Sub-typing with Ben Busby, DNAnexus

We’re happy to welcome back Ben Busby, Principal Scientist at DNAnexus, to dive deeper into the evolving world of disease subtyping and multi-omic data sets. Building on our previous conversation with Busby about the genomic data ecosystem, we explore the shift from single-cause disease models to multifaceted approaches that incorporate genomics, proteomics, imaging, and more.

"We’re no longer in a world where we’re doing genomics or imaging or proteomics," he explains. "These things are all coming together, and it’s important that they do. The UK Biobank will soon have 80 million images."

Busby highlights the UK Biobank’s leading role in this transformation, with its extensive genomic, imaging, and clinical data available to researchers worldwide. He emphasizes the importance of enabling scientists to access and analyze vast datasets collaboratively while ensuring participant privacy, and points to the role of DNAnexus in providing a trusted research environment where multiple stakeholders, from pharma to academia, can work with data efficiently and securely.

The episode also explores the impact of AI in bioinformatics, particularly in hypothesis generation. "AI is helping us think beyond single-cause events," Busby notes, referencing its ability to generate novel insights from complex biological data. He underscores the need for proper alignment between AI systems and scientific goals, stressing that human intuition remains essential in guiding these technologies toward meaningful discoveries.

As the field moves forward, Busby calls for more equitable data sharing practices, ensuring credit for data generators and benefits for study participants. "We need to figure out how to incentivize data generation in a way that’s fair and equitable," he says.

This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit www.mendelspod.com/subscribe