The Human Genome Project is one of the most significant scientific undertakings in recent decades. Its efforts to sequence all the genetic information within human cells led to groundbreaking biological discoveries and laid the foundation for an array of later research. This monumental effort, primarily funded by the National Institutes of Health (NIH) and the Department of Energy, continues to inspire new scientific endeavors. However, the role human-associated microbes play in human health remain poorly understood—specifically viruses associated with humans and their microbiome. 

The NIH recently funded an exciting new initiative to address this knowledge gap known as the Human Virome Program. This effort aims to explore the vast array of viruses that live in and on our bodies—collectively known as the virome—and uncover their impact on our health. Much of the existing human virome research focuses on viruses linked to specific diseases. This program takes a different approach by looking at viruses that coexist with us but aren’t tied to any known illnesses. 

Part of this newly launched program includes a $2.4 million project co-led by Pacific Northwest National Laboratory (PNNL) and the University of Arizona. Their efforts will focus on studying the interactions between viruses and the bacteria in our bodies, the latter of which makes up our microbiome. 

Combining strengths is the key to success

The interdisciplinary team on this new project is led by Jason McDermott from PNNL and Travis Wheeler from the University of Arizona. Together, they aim to accelerate the identification and characterization of viral sequences in the human microbiome to promote human health. This is bolstered by combining the lab’s strong capability in characterizing viral metagenome sequences with the university’s strengths in computational science.

“The incredible diversity of these viruses, coupled with our lack of knowledge, significantly limits our ability to understand, monitor, control, and potentially benefit from their interactions with bacterial communities found in our bodies,” McDermott explained. “Our project aims to illuminate this 'dark matter' by vastly improving our understanding of these viral elements.”

Over the next four years, McDermott, Wheeler, and their colleagues will develop open-source computational tools to better identify and characterize human microbiome viruses. Integrating artificial intelligence (AI) and machine learning (ML) approaches into new algorithms will also help the team handle large biological datasets more efficiently. 

Building upon past research

McDermott and Wheeler’s previous collaborations provide a strong foundation for hitting the ground running. 

“Typically, Jason and his team come to the table with significant biological challenges or questions,” said Wheeler. “Then my team uses our expertise in algorithm design, machine learning, and software engineering to implement solutions that address these challenges better than traditional methods.”

As their project kicks off, for the first year and a half the team is focused on building themselves a solid foundation that includes enhancing and augmenting their existing approaches to better handle virus detection and analysis. 

"The great news is that we already have a number of relevant tools at our disposal from past work researching soil viruses and other kinds of biological dark matter," McDermott said. "But they aren’t all directly applicable to viruses in the human microbiome. Our initial task is to considerably modify some, while fine-tuning others."

The research team plans to be flexible with their research priorities as their project progresses, particularly as they gain valuable input from a nationwide consortium of colleagues. This approach ensures that the project helps the broader scientific community as it fosters significant progress in biological and human health research.

The full project team also includes Ruonan Wu from PNNL and Jeremiah Gaiser, Clément Goubert, and Jack Roddy from the University of Arizona. This work is supported by the NIH’s Common Fund via grant number U01DE034176.