USC Computer Scientists Are Tackling Dental Health and Birth Defects

by Julia Cohen

Photo credit: USC

Roughly half of all birth defects involve the face and skull, yet scientists remain unclear about why most occur. 

The way to address tough medical challenges like this one is through data – lots of it. But how to best manage the data, integrate it into meaningful information, and create a comprehensive picture that is useful and accessible to researchers is another question. FaceBase offers an answer.

FaceBase is a research resource that provides open access to genetic, molecular and imaging data to the dental, oral and craniofacial (DOC) research community.

“Through FaceBase, USC is playing a role in the next generation of dental and craniofacial research,” said Carl Kesselman, FaceBase’s co-Principal Investigator, who is the William H. Keck Professor of Engineering and a Professor in the Daniel J. Epstein Department of Industrial and Systems Engineering and Director of the Informatics Systems Research Division at the Information Sciences Institute (ISI) in the USC Viterbi School of Engineering.  

He continued, “We are assembling all of the data, organizing the research community, and providing this service to the National Institute of Dental and Craniofacial Research [NIDCR] and the research community at large.”  

The Basics of FaceBase 

FaceBase is a collaborative NIDCR-funded project that houses comprehensive data in support of advancing research into craniofacial development and malformation.  

Kesselman, who is an ISI Fellow, leads the team of researchers and staff at ISI who run FaceBase’s coordinating center (i.e., the Hub).  

The Hub is where large datasets are curated and shared. Researchers in the DOC community can submit their projects to FaceBase, and datasets from approved projects are added. 

How does a large database help research? 

Rob Schuler, the technical lead for FaceBase and Senior Computer Scientist at ISI, gave examples of how some researchers are using FaceBase data: “to have a larger patient cohort; to compare their own clinical results with research being done on animal models; some of them do analysis and use the large datasets to train neural networks and produce models that can, for example, predict a phenotype based on a patient’s face.”  

But FaceBase is more than just an ever-growing database.  

FaceBase Connects the Dots 

“We don’t think of FaceBase as a data repository, although we do operate a repository as part of FaceBase. But really, we are an overall data resource,” said Schuler.

One of the missions of the project is to facilitate cooperation and collaboration between the Hub and the craniofacial research community.

“There’s a desire to be able to use a data resource like FaceBase to assist researchers in making connections to other people who are possibly working on a similar disease,” said Schuler. 

Kesselman said, “We connect the dots. In the absence of something like FaceBase, you have a little piece of data over here and a little piece of data over there, and you can’t figure out how they connect. But we do that. We take all these different aspects and research projects, we integrate them so that they’re more cohesive, and it represents more of the total knowledge of the community rather than isolated silos.”

The AADOC Conference

Kesselman and Schuler, along with Computer Scientist Alejandro Bugacov, Research Engineer Cris Williams, and USC Ostrow School of Dentistry Associate Dean of Research Yang Chai (Co-PI of FaceBase), recently made a big impact at the American Association of Dental and Craniofacial Research (AADOCR) conference in Portland, Oregon.  

From March 15 to 18, 2023, the team showcased their innovative work in the FaceBase platform’s data sharing and management. Bugacov presented a poster and provided demos at the NIDCR Trainee Research Presentation, which highlighted the platform’s user-friendly interface and powerful search capabilities. 

Meanwhile, Schuler presented two talks: an interactive talk on building FAIR data sharing communities (where he also served as session co-chair); and an invited talk in the Knowledge and Database Symposium.  

Big Praise From the Biggest Names in DOC Research 

The AADOCR conference also included a celebration of the 75th anniversary of NIDCR. 

The National Institute of Dental Research – which would later become the NIDCR – was founded as one of the earliest institutes of the National Institutes of Health (NIH); created in response to the tooth decay epidemic during World War II. At the time, oral health was an issue of national security, as potential military recruits were being disqualified from service due to tooth decay.  

Today, the mission of NIDCR is to “advance fundamental knowledge about DOC health and disease and translate these findings into prevention, early detection, and treatment strategies that improve overall health for all individuals and communities across the lifespan.” It has an annual budget of $520.20 million, funding approximately 770 grants, 6,500 researchers, 350 trainees and 200 organizations

With NIDCR being such a major player in the DOC research community, the FaceBase team was particularly excited to hear how their project is valued by the institute. During a panel, NICDR director Rena D’Souza called FaceBase an innovative data sharing platform; and current NIH director and former director of NIDCR Lawrence Tabak mentioned FaceBase as one of the top achievements during his tenure. Both endorsements speak volumes about the impact FaceBase has had in advancing the field of dental and craniofacial research. 

What’s Next for FaceBase? 

“We’re in the third phase of FaceBase right now, and we’ve opened it up to more projects,” said Williams. She continued, “Previously, we had specific ‘spoke’ projects around us, the ‘Hub.’ Ten to 12 projects were contributing data at a time. But now it’s open to the community, which has definitely widened the scope even further of what we’re taking in and how we’re building up our database.”

How wide of a scope? Kesselman gave examples of two current projects: “We’ve got a large dataset contributed by colleagues looking at the genetic foundations of tooth enamel. And another large set of data from researchers studying oral health in Appalachia. They’ve looked at social factors, along with all kinds of various health factors associated with oral health.” 

In addition to opening it up to more research, the FaceBase team is also looking at applications for clinicians, the people who are actually treating and diagnosing patients. Williams explained, “That’s not something FaceBase had focused on before. We’re working on a pilot project about what it would take to serve the clinician community, and that opens up a whole new frontier.”

Kesselman said, “We’ve been working with people in the Ostrow Dental School for the last eight years now, applying very sophisticated computer science and research that we’ve developed at ISI. And by applying it in this area we are ultimately making a real impact on dental health and childhood development.”  

About the FaceBase Co-PIs

Carl Kesselman is the William H. Keck Chair of Engineering in the USC Viterbi School of Engineering and is a Professor in the Daniel J. Epstein Department of Industrial and Systems Engineering. He also holds joint appointments as Professor in Computer Science at the USC Viterbi School of Engineering, the Department of Population and Public Health Sciences in the Keck School of Medicine and in the Herman Ostrow School of Dentistry. He is the director of the Informatics Systems Research Division at ISI and an ISI Fellow, the institute’s highest honor.

Yang Chai is a University Professor and he holds the George and MaryLou Boone Chair in Craniofacial Molecular Biology. He is the director of the Center for Craniofacial Molecular Biology and is Associate Dean of Research in the Herman Ostrow School of Dentistry.

Published on May 1st, 2023

Last updated on January 25th, 2024

Want to write about this story?