Machine learning to crunch genomic data in search of cancer cures for veterans

After two years of accumulating data from top-tier cancer research institutions, the IBM Watson for Genomics machine learning system is being put to the test solving complex cancer cases for US veterans. IBM is donating access to Watson for Genomics to Veterans Affairs (VA) hospitals nationwide, as part of Vice President Joe Biden’s Cancer Moonshot program. The VA estimated that it would be able to increase the number of patients who receive targeted therapy by 30-fold using the technology. The increase in caseload is possible because Watson can review an individual’s genomic data and incorporate the latest findings from clinical reports and medical literature within a few minutes, producing evidence-based recommendations on treatments that may be more likely to work with the individual’s unique DNA profile. The system is designed to learn continuously, to understand complex questions, and to respond in natural language, as demonstrated by its performance on the TV show Jeopardy! several years ago. Over the next two years, it is hoped that 10,000 veterans could be treated with targeted cancer therapies based on the insights provided by the Watson for Genomics system. Further, the collaboration could allow patients who do not live near major academic medical centers access to the insights and experience of physicians in those institutions to guide their treatment. —Karyn Hede, News Editor

China investing in genomic medicine on an industrial scale

While the National Institutes of Health was busy touting its Precision Medicine Initiative (PMI), China threw down the gauntlet with its identically named initiative, announced in March 2016. Whereas the United States is investing $215 million in funding for the first year, China announced an investment of US $9.2 billion over 15 years for its PMI. Working in parallel with the Chinese government, DNA-sequencing powerhouse Beijing Genomics Institute is expected to inaugurate the China National Genebank in Shenzhen by September, according to reporting in a special June 2016 issue of the journal Nature devoted to science in China. The Genebank will house samples and data not only of an expected 1 million human genomes but also genomes of plants, animals, and microbes. The government initiative includes investments not only in DNA sequencing technologies but, perhaps just as importantly, in information technology (IT) infrastructure. The China Precision Medicine Cloud will provide the IT infrastructure for organizing, mining, and sharing genomics big data throughout China and beyond. Far from being cloistered away, the Cloud meets international data standards for information exchange and is designed to be interoperable with technology used by the US Food and Drug Administration. The China initiative includes plans to sequence more than 1 million human genomes and then use that information, along with clinical and health data, to develop new drugs and diagnostics. Academic and industry partnerships, backed by massive government funding, could accelerate China’s ascendance to a world power in genomic research and its application to medicine. —Karyn Hede, News Editor