Scientists predict effectiveness of cancer treatment with computer algorithm

Scientists have developed a way to predict whether a particular therapy can kill cancer cells. They developed an algorithm that predicts the response of the cells, something that should help the development of new therapies.

The concept was developed by the university in Tel Aviv in collaboration with scientists from MIT and Harvard, among others. According to the makers, the developed algorithm works with so-called synthetic lethality. This means that changes that lead to inactivity in two or more genes together are lethal to the cell, while a change in a single gene is not. Identifying changes in gene pairs that are collectively synthetically lethal could help develop personalized cancer therapy by knocking out targeted genes.

According to the scientists, the development of synthetic lethality-based cancer therapy is hampered by the difficulty of figuring out which genetic changes are their ‘lethal partners’; after all, it is unlikely that inactive genes that jointly cause cell death are present in a cell at the same time. The algorithm developed by the university in Tel Aviv should change that.

The researchers examined a large dataset containing the genetic profiles of cancer cells. They adhered to the rule that genetically lethal partners, or inactivated genes, can occur individually in cells, but not together. This allowed them to detect genetic changes that are unlikely to be viable collectively. Based on the data, an algorithm was built that predicts whether genes are their lethal partners. Subsequently, a genetic network was drawn up from which the relationships between different genes became clear.

The developed algorithm was validated on the basis of predicting partner genes that were previously known to cause cell death. As a result, the scientists believe that the genetic network they have developed can help the development of new cancer therapies. The idea is that a patient is examined for which genes are inactive, after which a drug is used to disable his lethal partner, resulting in the death of the cancer cell.

This form of personalized therapy should increase the effectiveness of existing therapies and reduce side effects. Also, examining lethal changes in gene pairs may help unravel the biology underlying cancer development.