From W3C Wiki
Jump to: navigation, search

Business Cases


  • Therapeutic area
  • Class of drug
  • Competitive information

Evaluation Criteria

  • accessibly (effort, IP)
  • maintenance
  • coverage by others
  • participant interest/expertise
  • immedidate utility

Use Case Proposal

  • Define a use case that will allow us to interrogate the data from the perspectives of patients, pharma companies, and physicians.
  • Start the initial work focusing on the neuroscience area, but it should be possible to expand to any other therapeutic area.
  • It would be interesting if information relating to co-morbidity (disease overlap) is revealed.
  • Links would typically go from drug to disease to biology to symptom (genes-proteins-pathways-targets-diseases-drug-patient data)
  • Make the ClinicalTrials.gov data the center of the demo, as it incorporates high quality data that currently isn't linked, would appeal to patients, pharma companies, and physicians.
  • Patients are likely to want to ask questions such as:
    • What other drugs are available for this disease? - 1 (Don leads)
    • Are there ongoing clinical trials for this disease? - 1 (ClincalTrials.gov) (Scott, Susie leads)
    • Am I suitable for this clinical trial in terms of my location, ethnic background, and other inclusion criteria? - 1 (Bosse leads)
    • Are there natural alternatives to this drug? - 1 (DBPedia? - Jun, Kei & Huajun to investigate)
    • Are there contraindications between this drug and the others I'm taking (including alcohol, heavy equipment)? - 2
    • What side effects are there for this drug, especially those not on the label? - 2
    • What drugs work for a particular set of symptoms? - 2
    • What are people saying about this drug? - 2 (aggregation of data from blogs etc using SIOC)
    • Are there other people like me? - 3 (follow on to 'What are people saying about this drug?')
    • Are there simple changes that I can make to my lifestyle that may help? - 3
  • Physicians would like to know:
    • What drugs are available for a particular disease?
    • How do the drug compare in terms of safety, efficacy, and cost?
    • What food, drinks, other drugs, and other activities should the patient avoid while on this medication?
    • What lifestyle changes should I recommend?
    • How does the drug interact with the disease?
    • Is my patient a good candidate for a particular drug?
    • Are there any ongoing Clinical Trials that my patient could enroll in?
    • Is there interesting research or research results in this area (presented as a summary)?
    • Is there a combination of drugs that would work best?
    • (Possible link to the COI task?)
  • Pharma companies would want to ask
    • What other companies are running clinical trials for patients with similar symptoms?
    • What trial design are other pharma companies using for similar symptoms?
    • What other symptoms could our drug help with?
    • What other companies have drugs that are similar to ours in terms of compound similarity or therapeutic focus?
    • How similiar is the biological impact of closely related drugs?
    • What is the mechanism of action of our drug?

Pharma Business case

The pharma business case is built with the objective to demonstrate the value linking open drug data provide by allowing questions to be launched cross data domains. The business case is defined with generic questions and the most prioritized key concepts (identifiers) and data sources are identified.

  • Core principles:
    • Build from clincaltrials.gov.
    • Link based on key concepts, make sure key concepts are clearly defined.
    • Keep it simple, but make sure several different data domains is linked.

The pharma business case have the ambition to address needs from different domains, therefore typical clinical and discovery questions are defined to secure cross domain linking.

Questions Pharma companies may want to ask:

  • What other clinical trials is active for patients with similar indications?
    • Active trials?
    • What companies (sponsor)?
    • Where (investigators, address)?
    • What can be found for trials in social communities?
  • What trial design are used for similar trials? (#subjects,
    • Company?
    • Indication?
    • Phase?
    • Target?
    • Patient population?
  • For what other indications/symptoms could our drug help?
    • Known side effects?
    • Target
    • Mechanism of action
    • Indication
    • Pathway
  • What other drugs are similar in terms of compound structure or therapeutic focus?
    • Companies?
    • Target?
    • Side effects?
    • patient preference?
  • How similiar is the biological impact of closely related drugs?
  • What is the mechanism of action of our drug?
  • Concepts (Identifiers)
    • Compound
    • Target
    • indication
    • Company
  • Data sources
    • clinicaltrials.gov
    • drugbank.ca
    • www.ncbi.nlm.nih.gov/omim/
    • social networks?
    • Side effects?
    • Epidemiology data?

Pharma Use Case Steps 1

  • 1. Are there any particular regions in which prevalence of Alzheimer's Disease is high, as this may be a good location for holding clinical trials (WHO & Geonames)
  • 2. Are other trials being conducted in Massachusetts as that might influence the number of available patients for our trial (LinkedCT)
  • 3. The trial by Medivation looks interesting as it's using a drug that has been available in Russia for many years (http://clinicaltrials.gov/ct2/show/NCT00675623?term=Dimebon&rank=2&show_locs=Y#locn). Lets see details on the mechanism of action (MOA) and known side effects by linking to AlzForum (http://www.alzforum.org/drg/drc/detail.asp?id=111) using either the drug name (Dimebon) or ClinicalTrials.gov Identifier (NCT00675623)
  • 4. The MOA is described in the AlzForum as "Has activity as an inhibitor of cholinesterase and NMDA receptors. Inhibits neuronal death, potentially by mitochondrial-mediated inhibition of apoptosis", and side effects are described as "In a recent 6-month trial (in Russia, concluded in July 2006) Dimebon was well-tolerated. There were fewer serious adverse events in the Dimebon-treated patients than in the placebo-treated patients. The most common adverse event was dry mouth. All other gastrointestinal side effects were uncommon, occurring in less than 3 percent of patients (Medivation disclosure)".
  • 5. This looks like a good compound,I wonder what information may be available in Wikipedia (http://en.wikipedia.org/wiki/Dimebon).
  • 6. As wikipedia has a pubchem identifier (197033), lets take a look at the chemical properties too (http://pubchem.ncbi.nlm.nih.gov/summary/summary.cgi?cid=197033). Also known as Dimebolin
  • 7. Are there other uses for this medication?
  • 8. I've never heard of the company Medivation, lets see what they do (the company isn't in wikipedia, but maybe Open Calais could help?)

Pharma Use Case Steps 2

Disease ODS Disease XLS

Natural Alternative Use Case

Use case slides

This use case explores the use of social bookmarking and Semantic Web in helping patients find and share information about natural alternatives for disease treatment in a multilingual context. We use Alzheimer Disease (AD) as an example. For demonstrating use of multiple languages, we begin with English and Chinese. The social bookmarking tool we propose to use is Faviki (http://www.faviki.com/). Faviki allows users to use predefined semantic tags to annotate web pages. The source of these semantic tags is DBpedia (a Semantic Web resource). In this way, Web pages bookmarked by users (patients) are not just described by random words but connected to uniquely defined concepts that are semantically related. Faviki makes use of another dimension of semantic tags that has not been possible before — multilingual semantic tagging. It allows users to tag in 14 different languages, keeping Web resources connected to the main English version and translating tags into the users’ languages. This means that, for instance, a resource that is tagged by different users in Chinese and German tags, can be still found by using tags in Russian, because English is used as a universal reference.

Faviki is currently a case study of the W3C’s Semantic Web Activity (http://www.w3.org/2001/sw/sweo/public/UseCases/Faviki/). Our natural alternative use case aligns with the case study.

Web Resources for Alternative Treatment for Alzheimer’s Disease

There are Web resources for finding alternative treatments for Alzheimer’s Disease. Examples include Alzforum (http://www.alzforum.org) and Alzheimer’s organization (http://www.alz.org). Ginkgo biloba (a Chinese herb) is listed by these resources as one of the natural alternatives for treating AD.

MedlinePlus (http://medlineplus.gov/) is a patient-oriented, multiple language Web site developed by the National Library of Medicine. It has a section on drugs and supplements containing information about ginkgo biloba (see http://www.nlm.nih.gov/medlineplus/druginfo/natural/patient-ginkgo.html). For example, it provides safety information including allergies, side effects, and interactions with other drugs.

ClinicalTrials.gov (http://www.clinicaltrials.gov) is a registry of federally and privately supported clinical trials conducted in the United States and around the world. It includes clinical trials involving the use of ginkgo biloba in AD treatment. For example, one clinical trial (http://www.clinicaltrials.gov/ct2/show/NCT00042172) determines whether adding ginkgo biloba extract enhances the effects of donepezil. Donepezil is a drug that interacts with the cholinergic receptor family. For more detailed scientific studies of the molecular mechanisms underlying the drug action, one can search journal articles using PubMed. In addition, there is Web resource called “TCMGeneDIT” (http://tcm.lifescience.ntu.edu.tw/) that provides association information about herbs, genes and diseases. Such detailed information on mechanisms of action seems to fit with the pharma use case.

News articles represent another source of information patients keep track of in terms of medical information. For example, USA TODAY has recently published a news article (http://www.usatoday.com/news/health/2008-11-18-gingko-biloba-alzheimers_N.htm) on the results of a large clinical trial study on the use of ginkgo biloba in patients with AD. The conclusion was that ginkgo biloba doesn’t block AD.

Possible Intersection with BioRDF

Currently, the BioRDF group is exploring a query federation use case involving integration of receptor related information from multiple sources including DBpedia and HCLS KB’s. The natural alternative use case may intersect with the query federation use case. For example, it may be possible to integrate receptor information with Faviki’s bookmarks that are tagged with ginkgo biloba and receptor names that are found in DBpedia.