In the basement office of Jeff Hammerbacher at Mount Sinai’s Icahn School of Medicine, a supercomputer called Minerva named after the Roman goddess of wisdom and medicine was installed in 2013. In just a few months Minerva generated 300 million new calculations to support healthcare decisions. Dr. Joel Dudley, director of biomedical informatics at the Icahn School of Medicine, said that what they are trying to build is a learning healthcare system.
“We first need to collect the data on a large population of people and connect that to outcomes. Let’s throw in everything we think we know about biology and let’s just look at the raw measurements of how these things are moving within a large population. Eventually the data will tell us how biology is wired up.”
When they assembled and analyzed the health data of 30,000 patients who volunteered to share their information, it turned out that there might be new clusters or subtypes of diabetes. By analyzing huge amounts of data it might be possible to pinpoint genes that are unique to diabetes patients in these different clusters, providing potentially new ways to understand how our genomic background and environment are linked to the disease, its symptoms, and treatments.
Analyzing big data is key to the future of healthcare. But it’s not only about computational power, but a new paradigm about how we look at the networks of diseases. I loved the book, Burst, from Albert-László Barabási, the world-known expert of network medicine. It proved there are hidden patterns behind everything from e-mails to science.
I had a chance to meet him in person a few weeks ago and we chatted about his theories of network medicine for an hour. He thinks disease-disease relationships can be predicted and uncovered through the protein network, so-called interactome which is incomplete at this time. He and his team think that there are molecular fingerprints behind diseases and hidden structures which can only be uncovered with smart algorithms and bioinformatic methods.
The system they have been developing is aiming at interpreting gene expressions and genome-wide association study data to drug target identification and re-purposing. The name of Barabasi’s exciting start-up is DZZOM, derived from their map called „Diseasome”. Their approach and tools are certainly offering new opportunities to reclassify disease relationship from a network perspective and molecular level interactions. Obviously, biopharmaceutical companies are the primer targets for their services.
We will see how it transforms the way pharma companies develop new drugs and how it affects everyday medicine. Until then, read the paper published in Science.