In a proof-of-concept study, IBM’s Watson interpreted data from the whole-genome sequencing of a brain tumor more quickly than a team of experts. While the study is limited, the authors suggest it is an important step in scaling precision medicine for cancer.
IBM initially teamed up with the New York Genome Center in 2014 to create a version of its Watson supercomputer for cancer genomic research. In the study, researchers from the New York Genome Center, the Rockefeller University and IBM used a beta version of Watson for Genomics to interpret whole-genome sequencing from one patient with glioblastoma, according to a statement.
RELATED: NY Genome Center enlists IBM to create Watson-enabled cancer research platform
The study found whole-genome sequencing can reveal much more about a tumor than the current method, called a targeted panel, which only looks at a limited group of genes. But it can take a team of experts many hours to analyze these data and come up with treatment recommendations.
In the study, the researchers applied whole-genome sequencing to the glioblastoma tumor DNA and also sequenced the tumor’s RNA. They also analyzed DNA from the patient’s normal blood, they said. A team of oncologists and bioinformaticians took an estimated 160 hours to crunch the information. Watson processed the genomic data, as well as abstracts and full-text articles from PubMed, turning up “similar conclusions” in 10 minutes, according to the study, published in Neurology: Genetics.
RELATED: IBM, Illumina deploy Watson for Genomics in cancer research
“This study documents the strong potential of Watson for Genomics to help clinicians scale precision oncology more broadly,” said Vanessa Michelini, Watson for Genomics Innovation Leader at IBM Watson Health, in the statement. “Clinical and research leaders in cancer genomics are making tremendous progress towards bringing precision medicine to cancer patients, but genomic data interpretation is a significant obstacle, and that’s where Watson can help.”
“This study is an important step forward promoting human-machine interface as a way to address a key bottleneck in cancer genomics,” the authors said. Combining artificial intelligence with the whole-genome sequencing of tumor DNA could help doctors pinpoint potential targets and therapies for more patients, more efficiently. This is particularly important in patients with aggressive cancers, such as glioblastoma, where the five-year survival rate peaks at 17%.