The average scientific paper is, well, long. And
a new one is published every 30 seconds. Even the world’s best speed readers
can’t keep up. Thanks to the IBM Watson and IBM Research-Almaden teams, as well as partners like the Baylor
College of Medicine, Watson Discovery
Advisor can do all that page turning – learning and understanding along the way
– to help life science and pharmaceutical experts access and understand the
latest and most-relevant research related to their domain of expertise.
Because Watson can read millions of
documents in seconds, Discovery Advisor can cross-reference things like drug
interaction and disease symptoms – and return not just every piece of related
evidence, but visualizations that explain trends in the data, and offer
hypotheses for a researcher’s query.
with domain experts we have trained Watson to recognize some of the fundamental
concepts in life sciences research, such as genes, drugs, and diseases. Watson
can understand the various names that these entities go by, including a wide
range of international brand names and code numbers, so it can, in seconds,
retrieve all the relevant evidence in the literature for an entity of interest.
importantly, Watson understands the relationships between these entities by
reading and interpreting the literature. For example, it can rapidly return all
documentary evidence describing genes that regulate a disease like Alzheimer’s,
and summarize this information with intuitive visualizations.
can also reveal patterns in the evidence, and by comparing the entities make
predictions about potential connections that aren’t currently described in
literature. It’s constantly learning as we provide more and more expert data
and literature to ingest and cross-reference. Life sciences is just one of the
domains in which we’re already seeing huge potential for this kind of cognitive
– Richard Martin, Research
Scientist on the Watson Discovery Advisor team
IBM Research's Richard Martin (r) demonstrating Watson Discovery Advisor
There are more than 70,000 scientific
articles about the tumor-suppressing protein p53, often referred to as “the
guardian of the genome.” The p53 protein reacts to the detection of genomic
problems by increasing the expression of hundreds of other proteins to try and
fix these issues, and can even instruct potentially harmful cells to destroy
It gets these calls-to-arms from another set of proteins that
chemically modify p53 in response to particular biological conditions. But when
Baylor and IBM started work on the challenge of understanding p53, only 28 proteins
that perform that specific “phosphorylation” modification on p53 had been
discovered – over the last 35 years of scientific research. By putting Watson
to the task, it was just a matter of weeks before the teams had discovered and
validated five more potential candidates in laboratory experiments.