It’s Like Molecular Speed ​​Dating: LSU’s Use of AI in Cancer Treatments | education

A team of LSU researchers has developed a way to determine which drug treatments work best against an individual’s unique type of cancer, which could provide a way to find treatments more quickly and make treatment more affordable.

The multidisciplinary team includes researchers from the College of Veterinary Medicine, the College of Science, the College of Engineering, and the Center for Computation and Technology. You have created CancerOmicsNet, a new drug discovery engine powered by artificial intelligence.

Using algorithms originally designed to map complex social networks, such as those used by Facebook, researchers have created 3-D graphs of molecular datasets that include cancer cell lines, drug compounds, and interactions between proteins within the human body.

The graphs are then analyzed and linked together by artificial intelligence, forming a much clearer picture of how a particular cancer responds to a particular drug.

Dr. Michal Brilinsky, associate professor of computational biology at LSU, said the team used established datasets to train the CancerOmicsNet engine to use artificial intelligence.

“Once the training is done,” he said, “you can ask for something you don’t know and that’s the input data.” “So you ask what inhibitor you think would be effective against this cancer and then the AI ​​makes a prediction. That’s the implication of the unseen data and then something like that goes to the wet lab and we can validate it.”

Researchers at LSU College of Veterinary Medicine conducted wet lab research led by research associate professor Brent Stanfield.

“They developed the AI ​​algorithm and everything, so our role in the study is just to be the practical applications of the technology,” Stanfield said. “They developed the algorithm, they identified the drugs, and then we tested the drugs with our high capacity. To prove their effectiveness in killing cancer cells.”

Researchers studied highly aggressive breast, prostate and pancreatic cell lines to train the AI ​​to recognize the links between certain types of cancer and cancer drugs that control kinase production within the body.

Kinase acts as a biological catalyst for cellular communication and cell growth. Use of drugs that reduce kinase activity can suppress the growth of cancer cells.

Brylinsky said the research team used CancerOmicsNet to select and tested six groups of cancer cell lineages with the drugs likely to be the most toxic to their gene expression profile, with encouraging results.

“By accepted standards, four out of six people have succeeded and that success rate is very high because if you just pick six random drugs and say ‘these drugs will work on this cancer,’ it probably won’t work for this cancer,” he said of this cancer. “Four out of six were very encouraging,” he said. And that’s what we stand on now.”

Using CancerOmicsNet as molecular rapid dating, AI can help researchers quickly match cancer cell lines with drugs that are likely to be most toxic to their growth and genetic profile.

The knowledge gained through CancerOmicsNet could help overcome the challenge of determining the efficacy of a particular kinase-inhibiting drug in the future, Brylinsky said.

The ultimate goal, he said, is to expand the scope of their research for application in clinical settings.

“If we have a patient who has a specific cancer, they can take a biopsy and then they can identify that cancer in terms of gene expression, gene mutations, everything,” Brylinsky said. “Then they can enter this data into CancerOmicsNet and they can suggest some treatment for that particular cancer and say that this drug can be effective and ‘another drug that cannot be effective.'” “

It was initially thought that the effectiveness of different cancer drugs was linked to molecular consistency, and the idea that cancer treatment should target a specific site in a specific location in the body.

CancerOmicsNet is an example of how our current medical understanding of cancer treatment can meet advances in genetic studies and artificial intelligence, said Michelle Collins, dean of the Loyola University School of Nursing and Health in New Orleans and a scientist not involved in LSU research.

“When cancer drugs first came out, they were one-size-fits-all and weren’t really tailored to the individual and so you see that the drugs work better for some people than others,” she said. “With the advent of genetics and genomics, which is the future of medicine, we will now be able to design personalized treatments for the patient and not just in oncology.”

Collins said she sees CancerOmicsNet as very useful for future tumor studies and treatment.

She said, “I think it has the potential to really revolutionize the field of oncology, because we’ll be able to treat people with drugs that work for them at the right time. It’s all good if you’re sick. With cancer.”

Brylinksi said the ability to treat cancer with a more focused and focused clinical approach makes him excited to see how CancerOmicsNet has evolved over time.

“I don’t know if we’re going to make major advances in oncology anytime soon, but we’re making a cut where if enough people are doing it, the entire field is moving toward the goal of improving human health,” he said. “We’re very happy that we can make some contribution, which may not be a huge breakthrough down the road, but it’s definitely something beneficial for improving human health and that’s really cool.”

Leave a Comment