Health Tech Podcast: Using precision medicine to kill cancer — with artificial intelligence

Health Tech Podcast: Using precision medicine to kill cancer — with artificial intelligence

In December of 2012, Su-In Lee was working on the College of Washington in Seattle. In the future, she acquired the sort of name everybody fears: Her father, Cheol Lee, was identified with gallbladder cancer, a uncommon and aggressive type of the illness. By the point he was identified, his cancer had already unfold to different organs.

“It was a shock to our household,” Lee mentioned on the most recent episode of GeekWire’s Health Tech Podcast. “The subsequent day I simply went again to Korea and met his physician after which discovered that it’s incurable, particularly in that stage.”

Health Tech Podcast: Using precision medicine to kill cancer — with artificial intelligenceNot solely was his cancer incurable, there truly isn’t a single drug designed to deal with it. So Cheol’s medical doctors gave him a drug designed for pancreatic cancer, the organ subsequent to the gallbladder. They hoped it could give him a couple of extra weeks with his household.

“Whereas I used to be caring for my father, I stored considering that it could be actually nice if we will discover a drug that works for his cancer,” Lee mentioned. “After three months, he handed away.”

Lee was modified by her father’s passing. She got here again to Seattle with a brand new aim: Discover a approach to assist cancer sufferers get the very best remedy attainable, even when that solely means giving them somewhat bit extra time with their household.

Prior to now 5 years, Lee — a professor of laptop science and genome science on the UW — has grown that aim right into a precision medicine program known as MERGE. MERGE makes use of machine studying alongside with a affected person’s DNA and different well being information to predict which remedies will work greatest to assist them struggle the illness.

It’s simply one of many rising variety of precision medicine tasks which are leveraging an experience in artificial intelligence and a rising understanding of the human physique to struggle illnesses of all sizes and styles, from lethal cancer to the mysterious Alzheimer’s illness.

Hear Lee’s full story and be taught extra about precision medicine work in the latest episode of GeekWire’s Health Tech Podcast. Pay attention within the participant beneath or seek for “Health Tech” to subscribe in your favourite podcast app.

Lee was singularly geared up to meet the problem she set out for herself. She’s an professional in computational biology, utilizing superior laptop science and organic information to perceive the human physique.

“So I stored considering that if we perceive the genetic and the molecular profile of my father’s particular person cancer, we will doubtlessly discover the drug that’s going to work the very best for him,” Lee mentioned. “Even when it’s stage 4. For sufferers in that stage, it’s actually extending a number of months of the lifetime — it means so much, to the affected person and to the household.”

Heather Mefford, a pediatric neurologist and the deputy scientific director of Seattle’s Brotman Baty Institute for Precision Medicine, mentioned that the strategy isn’t nearly discovering the remedy that may lengthen the affected person’s lifetime. There are numerous causes precision medicine may very well be invaluable to a affected person.

To lower their threat for negative effects, for instance. To maximise the impact that the remedy that you simply give them goes to have at both treating or stopping a selected illness,” she mentioned. It’s all about “making healthcare as environment friendly as attainable for the person.”

The strategy can also be serving to scientists higher perceive the causes of some illnesses, like Alzheimer’s.

However Lee wished to deal with cancer sufferers. She teamed up with two Seattle researchers and hematologists: Dr. Pamela Becker and Dr. Tony Blau, whose startup All4Cure was featured on a previous Health Tech episode.

Blau and Becker are hematologists, or blood cancer consultants, so that they centered on understanding acute myeloid leukemia, AML. It’s one of the crucial frequent sorts of leukemia.

To coach the MERGE algorithm, Lee and her staff extracted genetic profiles from 30 blood samples taken from Blau and Becker’s previous sufferers. Thanks to new expertise, we will create genetic profiles like this in just below a day.

However the quantity of knowledge it creates is immense. One particular person’s genetic profile can maintain a terabyte of knowledge. In case you printed it into the average-sized paperback, that may be three million books, or about 20 p.c of the Library of Congress.

That’s the place machine studying is available in. Lee and her staff used the genetic information from these sufferers — alongside with information on the medicine they took and the way they responded — to practice a machine studying algorithm. Additionally they included publicly obtainable information on AML sufferers to increase the coaching set.

Because it sorted by way of the info, the algorithm discovered patterns and began to make connections between sure genes and medicines. Apparently sufficient, it made too many connections — Lee mentioned instructing the algorithm how to inform the distinction between vital and unimportant connections was truly difficult.

“If this gene does one thing vital in cancer, then that implies that this affiliation between that gene and in any medicine must be thought of vital,” Lee mentioned. Genes that regulate tear ducts, for instance, most likely aren’t as vital.

At the moment, MERGE is shut to precisely predicting responses to among the commonest AML medicine. Lee and her staff are persevering with their work and even engaged on a extra advanced model of the system that may predict responses to a number of medicine directly.

Their work has been constructed on the info of actual sufferers who had been keen to share it anonymously within the identify of science. Lee mentioned her work, and the work of others, depends on the generosity of nameless sufferers all around the world.

“I’ve been on each side,” Lee mentioned. “Sufferers actually ought to take into account giving their information. It’s like donating cash — it may be higher than that, truly. … I additionally acquired the info from my father’s pattern. It wasn’t truly straightforward, you recognize — being a affected person, I’m somewhat extra emotional than being a scientist.”

She shared the info with the hospital the place her father was handled. She wished them to use it for analysis, within the hopes that sooner or later it may lead to new remedies or cures for individuals like her father.

Editor’s observe: Researcher Suman Jayadev‘s work can also be included on this episode.

Related posts

Common Brain Malformation – Affecting About 1 in 100 Children – Traced to Its Genetic Roots


Unraveling the Origin of Alzheimer’s Disease – How the Disease Spreads in Human Brains


Global Health Innovators Mobilize to Help Developing Countries Combat COVID-19