San Francisco scientists using AI in hopes to better treat cancer

SAN FRANCISCO (KGO) -- Artificial intelligence is being used for vehicles, computer engineering and many other industries. Now scientists are also beginning to use AI to pinpoint the best ways to treat cancer.

"For the first time, we are going to be able to use artificial intelligence to be able to do millions, if not billions, of experiments. All in the computer, before we do the first experiment trying to kill cancer cells," said Dr. Deepak Srivastava, president of the Gladstone Institutes.

We got a look inside the Gladstone Institutes in San Francisco where teams are working together, using artificial intelligence to understand what experiments would be most impactful to kill cancer cells. Srivastava says they believe this approach will save lives.

"We will be able to take somebody's cancer, that so far has been untreatable, and be able to engineer their own immune cells in ways that it will specifically kill their cancer cell," Srivastava said.

MORE: Historic UCSF Asian American cancer study gets $12.45 million grant; participants needed

According to the American Cancer Society, skin cancer is by far the most common of all types of cancers in the U.S.

Colorectal cancer is the second most common cause of cancer deaths for both men and women. These two types are the main focus for this research.

"Part of what causes cancer is that our cells are failing to correct errors that occur randomly in our cells so there are many random errors in a cancer cell in a tumor. Only a few of them are actually driving the cancer, so our goal is to distinguish those from the random background," said Katie Pollard, director of the Gladstone Institute of Data Science and Biotechnology.

Pollard said their vision is to use machine learning models to figure out the patterns that distinguish cancer cell drivers.

MORE: Bay Area doctors using Ion robot to detect lung cancer safer, faster than before

"If we can do that then we could do much more effective experiments," Pollard said.

Pollard is leading this AI approach. Working fast to find ways to tackle cancer is personal to her.

"I'm a patient, I have patients in my family, and we don't have time to waste. I think that if we can go just a little bit faster, we will change people's lives," Pollard said.

Pollard said AI will help them figure out how to change T cells.

MORE: SF ultrasound clinic using FDA-approved AI to help detect types of breast cancer faster

"They would give a blood donation. They would have blood taken and have some of their immune cells taken out of their body. We would use CRISPR genome editing to change the cells and then put it back in the patient's body. Now with the ability to better kill cancer," Pollard said.

Typically, breakthroughs can take decades to make it to clinical trials. This team is aiming for a shorter timeline.

"My hope is that over a five-year period, we learned enough that we can design new types of clinical trials that can actually try to address those cancers that so far have been incurable," Srivastava said.

Gladstone Institutes received a $5 million grant from the Biswas Family Foundation to work on this research.