How researchers at UC Davis support the swine industry with data analytics on AWS

A research team led by Dr. Beatriz Martinez Lopez of the Center of Animal Data Modeling and Surveillance (CADMS) at the University of California, Davis, and supported by Amazon Web Services (AWS), received a prestigious National Science Foundation (NSF) Convergence Accelerator Phase II grant—one of only 10 such grants awarded. The grant funds research aimed at solving national societal challenges. While many NSF scholars tackle well-known issues like climate change and workforce development, Martinez-Lopez’s research is slightly less conventional: she studies swine health.

Why is this research groundbreaking? By 2050, the world will have to produce 70% more food to feed our growing global population, and meat demand alone will increase by 200%. Moreover, the swine industry is facing unprecedented threats from the African Swine Fever pandemic, endemic diseases, and emerging issues like antimicrobial resistance, that are costing producers millions of dollars per year. Without support, farmers will struggle to keep up—which is why Martinez-Lopez and her team at UC Davis, Iowa State University, and Carnegie Mellon University are collaborating with pig farmers, veterinary clinics, and other industry stakeholders to address these key swine health challenges.

Martinez-Lopez and her collaborators formed a convergent multidisciplinary team of experts, including computer scientists, epidemiologists, swine veterinarians and diagnosticians, software developers, data scientists, and bioinformaticians. The team is expanding and modernizing an existing CADMS data analytics platform, the Disease BioPortal, to help better prevent and manage diseases in swine populations worldwide. To make this optimized platform a reality, they are relying on scalable, cost-effective tools from the AWS Cloud.

Giving farmers data analytics tools to fight disease

For the last 10 years, Martinez-Lopez and her fellow researchers have been developing the Disease BioPortal platform, which aggregates and analyzes animal health data, to support decision making in animal health. This platform creates space-time-genomic advanced analytics and visualizations to help with outbreak investigations, risk assessments, trend identification, and more—but it was primarily designed for researchers and analysts. In the last few years, Martinez-Lopez’s group realized that to truly change animal disease management, they needed to expand the data and analytics integrated into the BioPortal and make that data more accessible and usable to veterinarians and farmers, not just researchers.

“We wanted to create a more intuitive and user-friendly interface that vets and farmers without data science skills could easily use in the field to early detect animal health problems or support risk-based preventive measures,” Professor Maria Clavijo said, one of the leading swine veterinarians and diagnosticians on the team. “We realized that if we can provide timely diagnostic information and insights about health risks and vulnerabilities of a farm to veterinarians and farmers, we can be more proactive about preventing diseases.” The ability to act swiftly is crucial when dealing with highly infectious diseases, such as African Swine Fever, which has an almost 100% mortality rate.

Disease BioPortal currently supports the nation’s top 10 swine producers. Collectively, these farmers produce 80% of the nation’s pork products. In its current iteration, the portal enables experts in the field to more quickly and simply conduct risk assessments, analyze trends, and improve farm management and biosecurity practices with benchmarking data. But Martinez-Lopez and her team aren’t stopping there. In this new NSF project, they are collaborating with cutting edge researchers specializing in artificial intelligence and data science to develop advanced prediction models and integrate key swine health data to identify patterns and detect potential threats before they get to the farm.

The Disease BioPortal team is now focusing their efforts on inputting, storing, and analyzing complex and diverse datasets from various diagnostic laboratories, farmers, and animal health organizations to create accessible analytical tools and user-friendly visualizations. They receive vast quantities of data each day from diverse sources—and quickly recognized that manually inputting this data was not scalable. To address this challenge, Zack Whedbee and Ron Williams, senior programmers for CADMS, started exploring solutions in the AWS Cloud. “UC Davis has been using AWS,” Williams explained, “so it made sense for us to explore AWS’s data analytics solutions. Eventually, we opted to migrate to a data lake because it’s flexible and scalable—which is exactly what we were looking for.”

The IT team will use AWS Glue to standardize and automate data collection from many different data sources as they migrate Disease BioPortal data into an AWS data lake. Amazon Athena is planned for data exploration and ad-hoc queries and AWS Glue for building data products that can be used by their current application. Once that’s accomplished, the team intends to leverage the data lake for predictive modeling using Amazon SageMaker. “Once we’re in the data lake, there are a lot of amazing things we can start exploring,” Martinez-Lopez noted. “We are interested in seeing how we can adapt artificial intelligence tools for health science, or text recognition, which will help us work toward the predictive modeling we want to build into the Disease BioPortal.”

AWS letter of support helps secure research grant win

AWS tools have been instrumental in the development and expansion of the Disease BioPortal, but that’s not the only way AWS is supporting this critical research. The AWS team supported Martinez-Lopez in the pre-proposal process for Phase II of the NSF Convergence Accelerator grant—a task that falls outside of the academic realm. “This proposal wasn’t easy for our researchers,” Martinez-Lopez said. “The NSF Convergence Accelerator is a more entrepreneurial program, so they don’t care that much about publishing papers. They want to see proof that we can create a solution that is scalable and sustainable after the grant is over. AWS helped us prove that Disease BioPortal was both scalable and sustainable.”

The AWS team helped the CADMS researchers with data management planning and provided a schematic for their data lake that helped demonstrate the platform’s viability. The AWS team also wrote a letter of support, a third-party testimonial included in the grant application, which Martinez-Lopez feels was critical to secure the grant funding. “With this proposal, we were trying to prove that we have the skills to solve this very important problem,” she explained. “And by working with the expertise of AWS researchers, the reliability and scalability of AWS tools and the support of a big company like AWS, we showed the committee that we are thinking long-term about this project—and have the support we need to bring it to life.”

Beyond swine: helping the livestock industry with cloud-based tools

Although the Disease BioPortal currently focuses mostly on swine health, Martinez-Lopez and her team hope to expand the technology to other industries in the future. “We started with the swine industry because they are the most progressive in terms of capturing and sharing animal health data, and there is a pressing need with the African Swine Fever  pandemic,” Martinez-Lopez explained. “But we have small projects with data from cattle, poultry and aquaculture and, as soon as we demonstrate the value of these new capabilities of the Disease BioPortal, we hope that other industries will start knocking on our door. Ideally, this system will eventually help the entire livestock industry, from dairies to poultry and beyond.”

“Not everyone in the industry is ready for data collection, standardization and implementation, which may be a challenge ahead,” Martinez-Lopez added. “But we know this technology has the power to transform livestock farming. By paving the way with swine, we hope that other farmers will see just how beneficial this tool can be to improve animal health, reduce production costs and guarantee a safe and reliable food supply.”

Learn more about AWS and how researchers are leveraging cloud technology at the AWS for Research and Technical Computing Hub.

Are you a researcher or principal investigator pursuing research funding? Request an AWS letter of support.

Read more stories about AWS for researchers:

• Introducing 10 minute cloud tutorials for research
• Preventing the next pandemic: How researchers analyze millions of genomic datasets with AWS
• Solving medical mysteries in the AWS Cloud: Medical data-sharing innovation through the Undiagnosed Diseases Network
• How MTI tracks social distancing efforts with the AWS Cloud and big data
• Cloud powers faster, greener, and more collaborative research, according to new IDC report
• How to set up Galaxy for research on AWS using Amazon Lightsail

Subscribe to the AWS Public Sector Blog newsletter to get the latest in AWS tools, solutions, and innovations from the public sector delivered to your inbox, or contact us.

Please take a few minutes to share insights regarding your experience with the AWS Public Sector Blog in this survey, and we’ll use feedback from the survey to create more content aligned with the preferences of our readers.