AI used to identify lung cancer vulnerabilities

The research team, led by scientists at the University of Montana, has focused on searching for vulnerabilities in lung cancer cells with innovative technologies.

The team used AI tools to improve their understanding of how protein groups in lung cancer cells regulate cell division and metabolism, with a view to develop better therapies to treat the disease.

“We examined how cells respond to anti-cancer drugs used to treat lung cancer,” said Mark Grimes, a University of Montana biologist. “We used machine-learning algorithms to detect patterns in data that are difficult to see because our human brains are not all that great at seeing patterns in large spreadsheets.”

Lung cancer continues to be a major cause of mortality. The disease is the third most common cancer in the UK, accounting for 13 per cent of all new cancer cases and affecting 48,500 new people every year, according to Cancer Research UK.

Although there are existing treatments for lung cancer, most of them are very invasive. Moreover, new drugs often work only for a limited amount of time, as cancer cells are able to evolve and adapt to the medications, creating new tumours. 

To solve this problem, scientists are looking to better understand the causes of the disease, to develop more effective treatment plans.

“The great thing about this work is that we’ve turned the patterns we discovered into networks that represent cell signalling pathways that are affected by cancer mutations and drugs that target mutated genes, called oncogenes,” Grimes said. “This work takes this approach to the next level by looking at interactions between the pathways, which are groups of proteins that work together in the cell.”

The use of AI tools gave the research team both higher-level and molecular-level views of the interactions between the pathways that cause cancer cells to divide and regulate their metabolism, according to the researcher. 

Cancerous tumours often have a hyperactive metabolism and limited supply of oxygen, which could be a target for new treatments.

“Identifying links between these pathways presents opportunities to attach vulnerabilities in the import and utilisation of nutrients in combination with other anti-cancer therapies,” Grimes said.

The finding of the research were published in the journal PLOS Computational Biology.