5737-B19 IBM Watson for Drug Discovery

IBM United States Sales Manual
Revised: June 9, 2020

Table of contents
Product life cycle datesProduct life cycle datesOperating environmentOperating environment
Program numberProgram numberPlanning informationPlanning information
HighlightsHighlightsSecurity, auditability, and controlSecurity, auditability, and control

Product life cycle dates

Top rule

Program Number VRM Announced Available Marketing Withdrawn Service Discontinued
5737-B19 00.00.00 2016-08-30 2016-08-31 2020-06-10 2020-06-10

Back to topBack to top

Program number

Top rule

  • IBM Watson for Drug Discovery (5737-B19)

Back to topBack to top


Top rule

Watson for Drug Discovery is a cloud-based, cognitive solution that provides dynamic visualizations and ranked predictions backed by passage-level evidence drawn from a wide set of heterogeneous public and private content, such as medical journal articles, textbooks, and patents.

IBM Watson for Drug Discovery empowers life sciences researchers to identify hidden patterns and connections as well as develop evidence- based predictive models from diverse, unstructured data sources at a scale and speed that is beyond what humans can do today.

Back to topBack to top


Top rule

IBM Watson for Drug Discovery core capabilities:

  • Aggregated, diverse content: Synthesizes massive public and published data sets, and has the ability to ingest private content (in a private instance).
  • Domain understanding: Knowledge of the terms of the language of life sciences, such as genes, drugs, diseases, and the relationships between them.
  • Cognitive technology: Leverages machine learning and natural language processing to detect connections and patterns that humans may not necessarily see. Organizations can apply predictive analytics to accelerate hypothesis generation and prioritization.
  • Agility and speed: Generates holistic network maps in real time to foster innovative research insights. Watson for Drug Discovery is always up-to-date and can quickly evaluate millions of pages of text through machine curation. Watson for Drug Discovery utilizes an adaptive and agile architecture that can change rapidly and iteratively.
  • Scalability: Combines infrastructure, big data, and machine learning at a scale that supports the needs of large enterprises.

Additional value for life sciences organizations:

  • Can help increase R&D efficiency: Drives more informed selection of potential therapy candidates and helps optimize portfolio management by enabling the organization to make more informed decisions regarding their portfolio
  • Helps accelerate insight generation: Identifies hidden patterns and connections as well as develops evidence-based predictive models from diverse, unstructured data sources at a scale difficult for humans to undertake
  • Can help increase researcher productivity: Accelerates the investigation of hypotheses and identification of novel ideas to augment productivity and output, creating potential for reduced time to market

Back to topBack to top


Top rule

Value of the cognitive computing capabilities of Watson for Drug Discovery:

  • Cognitive capabilities mean the knowledge base grows, allowing researchers to get more insights as the knowledge base grows.
  • Unlike a regular Google search that is based on keyword matching that returns a list of websites for the researcher to mine through on their own, Watson for Drug Discovery pulls out the relevant content and shows why that content is relevant.
  • As the technology continues to develop, the system can be trained to filter by more detailed and specific concepts.

Related concepts

  • Watson for Drug Discovery is more than keyword matching, it captures the different names of targets.
  • Watson for Drug Discovery understands relationships between entities and can understand different words that have the same meaning.

Predictive analytics

  • The challenge in drug research and discovery is finding more entities with a certain property. Watson for Drug Discovery goes beyond the existing training set to look at the content.
  • Reasoning analysis helps rank the entities closest to the set of entities already identified.

Public and private content

  • Watson for Drug Discovery pulls from 23 million Medline standard content source abstracts, over 3 million patents, and 700K plus full text journal articles.
  • Watson for Drug Discovery can potentially integrate organizations' private content, such as electronic lab notes, private ontologies, and so on.

Reviews content in an unbiased manner

  • Watson for Drug Discovery shows all relevant articles, allowing scientists to uncover commonalities they previously may not have considered.

Dynamic visualizations

  • Unlike similar products that only have static views of content, Watson for Drug Discovery allows researchers to interact with the views by filtering content differently and layering on and adding different entities.
  • Watson for Drug Discovery can filter visualizations dynamically based on fields,such as author, date, document text, gene, chemical, and drug, to name a few.
  • Researcher is able to see hidden connections and patterns.

Watson for Drug Discovery includes different visualizations to accelerate scientific discovery:

  • Entity Explorer: The Entity Explorer application can be used to search scientific literature to find documents that include user- specified entities (including synonyms), and display the documents with key entity types highlighted (for example, genes, drugs, diseases, and so on). The application also provides auxiliary displays, such as publication trends, journal names, Medical Subject Headings (MeSH) terms, and other metadata.
  • Co-Occurrence Table: The Co-Occurrence Table can be used to explore and discover affinities between different entities. Results are visualized as a table that shows the level of affinity between the entities in the query and other user-selected entities (for example, genes, MeSH terms, and so on) by virtue of their statistical co- occurrence in the source documents. The table helps reveal potential relationships between terms of different types, such as genes and diseases.
  • Biological Entity Network: Biological Entity Network visualization displays connections among genes, drugs, and diseases. These relationships are discovered by Watson for Drug Discovery using machine-learning annotators and they are not drawn from public structured databases. The user can click on individual links in the network to drill down into the supporting literature and understand how the relationship was discovered. Among other things, this application can help accelerate the identification of potential alternative applications for existing drugs and it can help to understand what biological pathways the entities of interest are involved in.
  • Reasoning Analysis: The Reasoning Analysis application is used to discover new entities related to a concept or entities of interest based on two query sets (that is, a training set and a candidate set, each specified by the user). The application helps users focus their hypotheses and discover new targets by mapping similarities between concepts or entities using semantic context. For example, if the training set consisted of proteins relevant to Type 2 Diabetes, the application would identify semantically similar proteins in the candidate set; such semantic similarities may suggest proteins that may also be relevant to Type 2 Diabetes. In this manner, this application helps the user to expand a set of research targets. This application reasons that queries are similar if they are surrounded by similar words and phrases. It computes a distance matrix that contains a similarity index for each pair of queries, and a predictive similarity score or heat index that measures each candidate's queries similarity to the training set. The higher the number, the more similar a query is to the entire training set and might be a research target to investigate further. Interactive visualizations allow the user to explore the similarities and the rationale behind them.
  • Chemical Search: The Chemical Search application can be used to find chemical compounds that are the same or structurally similar to a compound that the user specifies by name, chemical composition, or molecular structure.
  • Post-translational Modification (PTM) Summarisation: PTM Relationship Summarisation visualization can be used to learn about the post translational modification events for a specific protein. PTM is a step in protein biosynthesis in which important changes, such as changes to the chemical nature of an amino acid or changes to the structure and behavior of a protein, occur. Users can click on the interactive graph to get a list of scientific articles that describe the events that occur at that amino acid location and the agents involved in them. The PTM events are discovered by Watson for Drug Discovery from scientific articles using machine-learning annotators.

The production Watson for Drug Discovery environment is managed and hosted in the Watson Cloud Technology and Support cloud environment. However, the process initially starts in a development environment with a fixed set of public content on which the annotators run. The results are stored in an Apache Solr or IBM DB2 instance that Watson for Drug Discovery manages and tunes (as well as runs most of the performance testing) prior to pushing to Watson Cloud Technology and Support.

For a private instance of Watson for Drug Discovery, data storage and security are typically managed in the following ways:

  • Files are stored on an encrypted file system .
  • Private data files will be processed by Watson for Drug Discovery on the Watson Cloud Technology and Support servers.
  • Private data files will remain on the Watson Cloud Technology and Support servers and not be moved or copied from the Watson Cloud Technology and Support servers without prior consent from the organization.

Private instance of Watson for Drug Discovery consists of a private VLAN for access with dedicated hardware, management tools, databases, cognitive services, and application instance and tooling. Encryption is enabled end to end, including data at rest.

Accessibility by people with disabilities

A US Section 508 Voluntary Product Accessibility Template (VPAT) containing details on accessibility compliance can be found on the IBM Accessibility website.

Back to topBack to top

Operating environment

Top rule

Hardware requirements

Internet connection is required.

Software requirements

Browser is required.

Back to topBack to top

Planning information

Top rule

Customer responsibilities

The customer is responsible for evaluation, selection, and implementation of security features, administrative procedures, and appropriate controls in application systems and communication facilities.


For additional information, refer to the license information document that is available on the IBMSoftware License Agreement website.

Back to topBack to top


Top rule

Product documentation is available in IBM Knowledge Center.

Product documentation is available only in English.

IBM Knowledge Center is the repository for IBM product documentation. Customize IBM Knowledge Center to design the experience that you want with the technology, products, and versions that you use.

The documentation contains information for business analysts wishing to understand the product functionality, for developers wishing to understand all aspects of developing with the product, and for administrators involved in administration and configuration of the product.

Back to topBack to top

Security, auditability, and control

Top rule

Logical security is comprised of several technical controls. The following technical controls are in place:

  • Activity logging that includes suspicious activity monitoring of protected logs.
  • End-to-end encryption of protected or regulated data.
  • Isolation of protected or regulated data. Procedures for an emergency shutdown to prevent data leakage.
  • Technical specifications that detail allowable configurations for all devices.
  • Timely application of security patches.
  • Network configuration that includes zoned security layering enforced by mandatory firewall and router rule sets.
  • Security measures for user workstations.
  • Antivirus and anti-malware protection with automated workstation compliance tools.
  • Intrusion detection systems.
  • Change management process and information systems maintenance.
  • Audit logs recording privileged user access activities, authorized and unauthorized access attempts, system exceptions, and information security events shall be retained, complying with applicable policies and regulations.
  • Audit logs that contain sufficient information to, at a minimum, establish what type of event occurred, when (date and time) the event occurred, where the event occurred, the source of the event, the outcome (success or failure) of the event, and the identity of any user and subject associated with the event.
  • When audit logs contain sensitive data and personally identifiable information appropriate privacy protection measures must be taken.

Watson back-end systems use native identity, policy, and audit enforcement controls in conjunction with IBM Tivoli Identity Manager workflow and secondary controls providing an identity and access control system with robust security attributes used to manage access for Watson system administrators. Access to systems that host Watson software as a service (SaaS) offerings is approved based on role requirements. The appropriateness of a user's access is determined using the principles of least privilege and separation of duties as guidelines. Tivoli Identity Manager is used to retain audit trails of information related to the access control workflow, such as approvals. A periodic revalidation of user access, based on continuous business need and employment verification, is also performed.

IBM employs the latest cryptographic technologies and employs end- to-end encryption. Technology examples include Transport Layer Security (TLS), Internet Protocol Security (IPsec), public key infrastructure (PKI), third-party certification authorities, encrypted file systems, encrypted operating systems, encrypted storage systems, IBM Security Key Lifecycle Manager (ISKLM), encrypted application, among others.

IBM uses the latest tools for encryption key management.

Back to topBack to top


Top rule

(R), (TM), * Trademark or registered trademark of International Business Machines Corporation.

** Company, product, or service name may be a trademark or service mark of others.

Terms of use

IBM products and services which are announced and available in your country can be ordered under the applicable standard agreements, terms, conditions, and prices in effect at the time. IBM reserves the right to modify or withdraw this Sales Manual at any time without notice. This Sales Manual is provided for your information only. Additional terms of use are located at

Terms of use

© IBM Corporation 2020.