Frequently asked questions

Get answers to the most commonly asked questions about this product.


Getting started with this product

What is Watson for Drug Discovery?

Watson For Drug Discovery is made for researchers in the information-intensive pharmaceutical industry who are responsible for formulating conclusions in areas where consensus answers do not currently exist. It analyzes millions of documents and synthesizes multiple sources of information into a summary set of hypotheses with visibility into supporting evidence.

What are some common use cases that Watson for Drug Discovery supports?

Examples include:

  • Target Identification, such as identifying biological pathways of conditions or identifying new genetic markers or targets from literature through understanding context, entities, and relationships.
  • Gene function/regulation prediction: for example, uncover new potential roles for genes and proteins.
  • Drug insights, such as identifying alternative indications for a selected drug based on evidence from literature or identifying alternative compounds for a selected indication based on evidence from literature.

What are the key capabilities in Watson for Drug Discovery?

The key components of Watson for Drug Discovery are the deep natural language processing (NLP), domain knowledge, predictive analytics, and machine learning.

  • These capabilities can accelerate research with confidence through dynamic visualizations backed by scientific literature. For example, for drug target identification, Watson for Drug Discovery helps researchers see possible new relationships between genes, diseases, drugs and potential targets.
  • For drug questions, Watson for Drug Discovery helps researchers find potential relationships between shared disease pathways and shared drug mechanisms of action.

What types of research questions can Watson for Drug Discovery help answer?

The system is best used when the question asks what rather than why. For instance, what genes/proteins, drugs, diseases to target rather than why a drug works. And, it helps if you know positive examples of what is being sought (either “known” or “current best theories”).

Does Watson for Drug Discovery come up with answers for new targets or drugs?

Watson for Drug Discovery is not designed to present a user with a single, best answer to a complex drug discovery question. Rather, the system helps researchers filter through thousands of potential hypotheses in a systematic way to prioritize research directions most likely to yield positive results, making the researcher’s tasks less arduous and more efficient.

Does IBM have access to our saved results and entity sets?

The saved results and entity sets are stored within the Watson for Drug Discovery system so that they remain available to the user who created them - and to no other users. This data is automatically transferred into the new release of Watson for Drug Discovery every three weeks so that users do not lose access to any prior work. IBM staff do not access the saved results and entity sets. The saved results and entity sets are not examined nor analyzed by IBM in any way.