UCLA leverages Amazon Web Services to enhance computational medicine research

Apr 25, 2019

By UCLA Samueli Newsroom

A kickoff scientific symposium highlighted some of the collaborative projects in computational medicine using machine learning

Major advances in big data, artificial intelligence and cloud computing combined with a proliferation of biomedical data have opened up new possibilities to find insights in human disease. New knowledge from such efforts could lead to new therapies for diseases.  However, these discoveries require expertise in both computation and biomedical sciences. 

At UCLA, the David Geffen School of Medicine at UCLA and the UCLA Samueli School of Engineering have formed a collaborative effort, with faculty from both schools working together in this emerging area of computational medicine.

One particular focus area under this umbrella is computational genomics – which uses statistical techniques to understand the underlying genetic causes of diseases. More than a dozen UCLA researchers are looking to analyze that data using machine learning and data science techniques, with the aim to then translate that knowledge into effective patient care, and eventually, into cures.

While data science in medicine is nothing new, the scale of available data presents new challenges and opportunities as datasets with the anonymized genomes of millions of individuals are now available.  Obtaining insights from so many human genomes requires powerful tools to sort through and make sense of the mountains of information that it provides. To do this, UCLA has collaborated with Amazon Web Services (AWS), which will provide machine learning and cloud computing technology and expertise for these research projects.

“This type of research is much more complicated than say, finding a needle in a haystack,” said Eleazar Eskin, professor of computer science and human genetics at UCLA. “What we’re doing with our projects in computational genomics is more like looking at millions of haystacks and figuring out why just a handful of bad needles are in the spot they’re in. Then after that, we figure out effective ways to counteract the effects of needles, and even removing them entirely. To do that, we’re really going to need both dedicated computing power and expertise. AWS is the perfect cloud provider in that regard.”

“Machine learning is becoming a major catalyst in accelerating medical research, and at AWS, we are eager to put machine learning into the hands of researchers through a broad and deep set of services and powerful computational infrastructure in the cloud. We are excited to bring our machine learning technology and expertise to a wide variety of research projects ranging from cancer research and genomics to solving depression as part of our work with UCLA,” said Swami Sivasubramanian, vice president of machine learning at Amazon Web Services.

To kick off this collaboration, UCLA and AWS hosted a symposium on Computational Medicine on Feb. 1. The investigators leading some of those projects presented their work, including how the brain makes and keeps memories, what the genetic causes of depression are, and what role ancestry plays in evolution and disease.


Videos of the presentations from the Feb. 1 scientific symposium.
Opening remarks by Eleazar Eskin and Dean Jayathi Murthy

Jennifer Listgarten • UC Berkeley
Machine learning for protein engineering

Jonathan Flint • UCLA
Genetics of Depression

Ben Raphael • Princeton University
Clonal evolution in tumors and metastases

Share this article