Emma Hinchliffe

Watson consumed all published literature related to ALS, and learned all the proteins already known to be linked to the disease.
The computing system then ranked the nearly 1,500 genes in the human genome and predicted which could be associated with ALS. Barrow’s research team examined Watson’s predictions, and found that eight of the 10 genes proposed by the computer were linked to the disease. Of those, five had never before been associated with ALS.

One thought on “Emma Hinchliffe

  1. shinichi Post author

    IBM’s Watson makes its first discovery in neuroscience

    by Emma Hinchliffe

    http://mashable.com/2016/12/14/ibm-watson-als-research/#MjmLT_fqgGq4

    IBM Watson is known for its work in identifying cancer treatments and beating contestants on Jeopardy! But now the computing system has expertise in a new area of research: neuroscience.

    Watson discovered five genes linked to ALS, sometimes called Lou Gehrig’s disease, IBM announced on Wednesday. The tech company worked with researchers at the Barrow Neurological Institute in Phoenix, Arizona. The discovery is Watson’s first in any type of neuroscience, and suggests that Watson could make discoveries in research of other neurological diseases.

    “There are a lot of opportunities if we’re working together,” Robert Bowser, the doctor who led Watson’s ALS research as head of Barrow’s Gregory W. Fulton ALS Research Center, told Mashable. “There’s more room for Watson to grow its knowledge base in neuroscience. We’re hoping it’ll contribute to the process as Watson gains more and more information about neuroscience and the ability to work in the most complex organ in the body.”

    ALS, or amyotrophic lateral sclerosis, is a degenerative disease in which cells that control voluntary muscle movements die. That leads to paralysis and, eventually, death. The disease has no known cause, and there’s just one FDA-approved medication that only slightly eases symptoms. The disease gained major attention in 2014 with the viral Ice Bucket challenge.

    Watson consumed all published literature related to ALS, and learned all the proteins already known to be linked to the disease.

    The computing system then ranked the nearly 1,500 genes in the human genome and predicted which could be associated with ALS. Barrow’s research team examined Watson’s predictions, and found that eight of the 10 genes proposed by the computer were linked to the disease. Of those, five had never before been associated with ALS.

    The discovery gives researchers information about what to target for therapy when developing drugs to treat ALS. The program involved in this research is called Watson for Drug Discovery.

    Bowser said his team chose to use Watson in this way because of discoveries made in the past decade surrounding links between ALS and genes for RNA-binding proteins, or the proteins connected to the genetic material RNA.

    The research process took months, compared to the years it might have taken traditional researchers, who wouldn’t be able to read everything ever published about ALS.

    “We were interested to see if we could use Watson to accelerate the pace of discoveries,” Bowser said.

    Watson has been used in oncology research, including through Vice President Joe Biden’s Cancer Moonshot initiative.

    But this discovery in neuroscience suggests further potential for Watson in other fields, Bowser said. After the supercomputer’s successful work on ALS, researchers will likely target other neurodegenerative diseases, such as Alzheimer’s and Parkinson’s diseases.

    Researchers could also apply Watson’s abilities to neuropsychiatric and neurodevelopmental diseases and conditions, including autism, schizophrenia and bipolar disorder.

    For ALS, the next step is for Watson to explore remaining questions, like why the disease looks different from patient to patient. The computer will analyze data collected and produced by human researchers at the Barrow Neurological Institute.

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *