Scientists in Singapore have made a unique discovery about how to treat cancers – when it comes to pinpointing cancer treatment targets, it is better to listen to many computer programmes rather than just one. Researchers have developed an advanced system that integrates this ‘wisdom of the crowd’ through a powerful consensus algorithm to isolate the Achilles heel of each individual cancer tumor, helping scientists to better study different cancers and identify targeted treatments.
Cancer cells have thousands of genetic lesions but only a handful of these mutations give rise to a tumor. Identifying the ‘driver’ mutations that promote the uncontrolled growth of cancer cells in the body is a key challenge for the emerging field of precision oncology. This is the first time that scientists have identified a consensus algorithm that integrates various expert systems into a single accurate prediction for treatment targets in individual cancers.
To develop this system, researchers analyzed data from more than 3,000 tumors, across 15 different cancer types including colon, breast, lung, stomach and liver cancer. They studied 18 different existing algorithms and found that each one of them on its own could not identify driver mutations in a significant proportion of patients. No single method was able to identify treatable drivers in more than 60 percent of patients. The new system, known as ConsensusDriver, was able to identify treatment targets in nearly all patients studied, 80 percent of whom could be treated with existing drugs.
Cancer is among the leading causes of death worldwide and was responsible for 8.8 million deaths in 2015, and an estimated global economic impact of approximately US$ 1.16 trillion in 2010. Over the years, breakthroughs in DNA sequencing technologies have allowed researchers to determine the complete genetic makeup of cancers. The challenge now lies in crunching massive datasets to understand the unique genetic basis of an individual’s disease. Researchers around the world and in Singapore are now racing to develop new computer algorithms, and participate in large collaborative projects such as The Cancer Genome Atlas.