HCLSIG/PharmaOntology/UseCases
Use Cases
Patient Scenario
- 1. Patient [patient role OBI:0000093] (and family members [NCI: Patient_Family_Member_or_Friend CL357040]) report symptoms [IDO:0000048 Symptom] to physician / clinician [NCI Thesaurus: Physician]. Physician / clinician enters reported symptoms into eHR.
- 2. Physician [NCI Thesaurus: Physician] makes a list of differential diagnoses, with a working diagnosis [OBI:0000075] of Alzheimer Disease [DOID: 10652]. (Data Source: Physicians head)
- 3. Physician [NCI Thesaurus: Physician] arranges for patient [patient role OBI:0000093] to have a basic biochemical/haematological, and SNP [SO:0000694 SNP] profile undertaken. Biochemistry, Haematology, and SNP requests are input by respective departments directly into patient’s eHR HL7:EHR (UMLS C1555708 HID 20081) from laboratory (Data Source: eRecord). Preliminary SNP and genetic data will be submitted directly to the NIH Pharmacogenetics Research Network (PGRN).
- 4. A follow up meeting is scheduled to perform a set of diagnostic tests [NCI Thesaurus Diagnostic_Procedure] outlined by what the clinician [NCI Thesaurus: Physician] feels initially are most appropriate for disease presentation [obi:OBI_0000155]. (Data Source: Physicians head)
- 5. Physician [NCI Thesaurus: Physician] goes to ‘interface’ and continues to add investigations/lab results and these are combined with the patient’s medical history (occupational exposure, concurrent medication, lifestyle information) (NCI Thesaurus Personal_Medical_History), and a disease (AD [DOID: 10652]) is chosen as the most likely of the listed differential diagnoses (MeSH D003937) based on all of the information provided. (This is real time information). (Data Source: The National Institute of Neurological and Communicative Disorders and Stroke, and the Alzheimer's Disease and Related Disorders Association (now known as the Alzheimer's Association); and The Diagnostic and Statistical Manual of Mental Disorders DSM-V-TR criteria published by the American Psychiatric Association; Harrison's online, NICE, The National Guideline Clearinghouse) - use ICD9 codes
- 6. Physician [NCI Thesaurus: Physician] confirms and now has a refined and widely acceptable diagnosis of AD [DOID: 10652] with behavioural assessments, cognitive tests, and appropriate brain scan if indicated and enters data into the patient’s eHR HL7:EHR (UMLS C1555708 HID 20081). ICD
- 7. Physician [NCI Thesaurus: Physician] goes to the ‘interface’ to select the most appropriate AD drug [efo:EFO_0001899] & clinical protocol from the patient's medical record based on the severity of the disease, patient’s SNP profile (ADME, efficacy/safety (based on presence or absence of receptors)), patient’s BMI, availability on Medicare D, and concurrent medication.
- Fundamental questions will be answered by the ontology at this stage by sourcing the data sets listed below simultaneously or in a specific order.
- A. What are the clinically recommended agents NICE
- B. What products are available to prescribe, and which are legally indicated for disease AD [DOID: 10652]? [Physicians Desk Reference]
- C. What is the SNP verdict? These agents are sourced with pharmacogenomics database Pharmacogenetics Research Network (PGRN) to determine (Data source: PharmaGKB, Michel's DB (depression))
- Will they be efficacious? Receptor positive disease?
- Will they be harmful? Toxic metabolites? Available CYP 450 or acetylator status? (Data source: Drugbank, Dailymed?, SIDER?)
- D. Are the resulting pharmaceutical agents covered by the patient’s specific insurance? (In real time). Are the preceding predictive genetic SNP tests covered by the patient’s insurance company? These may be recommended and indicated prior to treatment as in HIV medication ABACAVIR.(PGRN, PDR, NICE, Medicare, Blue Cross Blue Shield)
- Fundamental questions will be answered by the ontology at this stage by sourcing the data sets listed below simultaneously or in a specific order.
- 8. The Physician [NCI Thesaurus: Physician] checks with pharmacist, or consults drug information literature to avoid potential drug interactions Facts and Comparisons
- 9. Physician [NCI Thesaurus: Physician] now prescribes Aricept (Donepezil) as it satisfies criteria A through D above. It is indicated, safe, effective, available, there are no drug interactions issues with drug delivery, and most importantly, it is covered.
- 10. In follow up, patient [patient role OBI:0000093] later reports nausea from donepezil, and Physician [NCI Thesaurus: Physician] is aware of this common side effect (other side effects reported include bradycardia, diarrhea, anorexia, abdominal pain, and vivid dreams etc…) re-consults literature to ensure this is acceptable and agreeable with patient [patient role OBI:0000093]. If not, revisit loop above. Document side effect for post marketing adverse event pick up MedWatch, and future study. Change medication if necessary or add another medication to alleviate side effects. Micromedex, Facts and Comparisons. Consider moving patient to a trial.
- 11. Physician opens ‘interface’ to view all trials for AD. These are local, national, and international. (Data sources: FDA, WHO, ClinicalTrials.gov, Citeline TrialTrove) (note academic or pharma may solicit patient, or physician may point patient to investigator undertaking the trial)
- Decision to :
- A. Enroll patient in a clinical trial as one of the agents looks very suitable and may benefit patient, or patient interested
- B. Do not enroll as patient declines, or trials unsuitable
- C. Obtain information for patient, with potential of future enrollment or not depending on best interest of patient.
- Decision to :
- 12. Physician checks patient meets inclusion/exclusion criteria through querying their eHR
- 13. Patient has thorough medical (lifestyle assessment, medical history, genomics, proteomics, metabolomics, images, cognition) to supplement and update existing data.
- 14. Results of the medical exam influence the arm of the trial in which the patient participates. Patient status updated.
More details on the data sources.
Research Scenario
- GWAS study [TMO:0029 study] in Nature genetics implicates several other proteins includiung CLU, PICALM, CR1.
- Virginia Lee [TMO:0009 expert] presents at a conference the hypothesis [OBI:0000074 hypothesis] that TAR DNA binding protein, TDP-43, may be an important factor
- Investigation of [TMO:0029 study] of gene [SO:0000704 gene], protein [PRO:000000001 protein], pathway [TMO:0044 pathway], and mechanisms of action [TMO:0042 MOA] to identify known druggable targets [TMO:0006 target] (Data Source: Drugbank) and potential side effects [TMO:0043 adverse drug event] (Data Sources: Entrez Gene, UniProt, KEGG, SIDER?)
- Identify compounds [CHEBI:23367 molecular entity] known to interact with that target [TMO:0006 target], including compound families and fragments
- Explore whether existing drugs [TMO:0001 pharmaceutical formulation] on market [TMO:0028 market] or in trials [TMO:0032 clinical trial] interact with the newly implicated pathways [TMO:0044 pathway] (Data Source: DrugBank)
- Determine prevalence of SNP [SO:0000694 SNP] of patients [OBI:0000093 patient role] enrolled in current trials [TMO:0032 clinical trial]. Consider recruiting more patients [OBI:0000093 patient role] to a new arm [TMO:0046 arm of clinical study].
- Segment patients [OBI:0000093 patient role] according to those that exhibited a good response, no response or an adverse effect.
- Determine significance of genetic variation to drug response according to clinical protocol
- Stratify or design a new study using personalized data in eHR with patient’s permission
- Use SNP genomic data to
- A. Perfect or focus treatment with agents already in clinical use to make them more patient specific, to avoid harmful effects, or ensure they are more efficacious
- B. Use data in Phase I for predictive toxicity study and analysis
- C. Use data/receptor status to identify additional targets for drug development
- Medical information [TMO:0020 medical history] from patients [OBI:0000093 patient role] involved in drug (Varenicline) [CHEBI:23367 molecular entity] trial [TMO:0032 clinical trial] is sent to researcher [TMO:0009 expert] in pharma [TMO:0025 institution] to further investigate the genetic underpinnings of the disease [TMO:0047 disease]
- Researcher [TMO:0009 expert] compares genetic variations such as single nucleotide polymorphisms [SO:0000694 SNP] from patients [OBI:0000093 patient role] and confirms that variations cluster on APOE [uniprot:P02649, PRO:000000001 protein], and may be significant [TMO:0017 piece of evidence].
Old Use Cases
Alzheimer Disease Therapeutic Development
- What are the current hot topic areas of research in Alzheimer Disease? (http://hypothesis.alzforum.org). First hot topic is “Abeta accumulation in the brain is the primary event in Alzheimer Disease pathogenesis”. Found two hypotheses, Hardy-Selkoe 2002 outlines the Amyloid Hypothesis, and Golde 2005 discusses therapeutic strategies based on the Amyloid Hypothesis. It identifies gamma secretase as an active area of research and development of AD therapeutics.
- Search keyword “gamma secretase” in Alzforum Drugs in Clinical Trial database (http://www.alzforum.org/drg/drc/default.asp?type=keyword&keyword=gamma+secretase) found 3 results. Two drugs discontinued, and one, LY450139 dihydrate (http://www.alzforum.org/drg/drc/detail.asp?id=108) currently in phase III of clinical trial. This entry also links to Alzforum news articles (http://www.alzforum.org/new/detail.asp?id=1607 http://www.alzforum.org/new/detail.asp?id=1992) which provide a historical perspective as well as the latest research and development of this drug. The news articles are linked to comments from the Alzforum community. One of the comments is from a family member of a patient sharing their experience in one of the trials (http://www.alzforum.org/new/detail.asp?id=1992#{5EB8EA74-C375-43CC-956D-D24854263ABE})
- Search clinicaltrial.gov for LY450139 (http://www.clinicaltrial.gov/ct2/results?term=ly450139) yields four results, with two trials actively recruiting. Physicians and patients will be able to identify study locations for the trials.
- Search ‘gamma secretase inhibitor’ in the USPTO Patent Database, (http://patft.uspto.gov/netahtml/PTO/search-bool.html) yields 57 documents. There are many valuable references and lists of inventors in this result set who are engaged in therapeutic development of gamma secretase inhibitors. These references and names could be circled back to look for relevant literatures.
- Pharma companies could link these resources back to their internal project management systems.
Chemogenomics Use Case
Chemogenomics tries to fill existing holes by predicting compounds–genes/proteins relationships.
Chemogenomics combines genomics questions with chemoinformatics questions. Ideally one would like to screen all possible molecules against all possible targets, but the search space of the former is practically infinite, so the focus is on _families_ of molecules against _families_ of targets.
Chemogenomics will be at the interface of chemistry, biology and consequently informatics since data mining is required to extract reliable information.
Chemogenomics is the new interdisciplinary field, which attempts to fully match target and ligand space, and ultimately identify all ligands of all targets (Caron et al., 2001). Methodologies at the border of chemistry and biology (medicinal chemistry), chemistry and informatics (chemoinformatics), biology and informatics (bioinformatics) will also play a major role in bringing these major disciplines together. (Rognan BJ Pharm 2007)
Chemogenomic approaches to drug discovery rely on at least three components, each necessitating hard experimental work: (1) a compound library, (2) a representative biological system (target library, single cell and whole organism), and (3) a reliable readout (for example, gene/protein expression, high-throughput binding or functional assay) (Rognan BJ Pharm 2007)
Analysing chemogenomic data is a never-ending learning process aimed at completing a two-dimensional (2-D) matrix, where targets/genes are usually reported as columns and compounds as rows, and where reported values are usually binding constants (Ki, IC50) or functional effects (for example, EC50).(Rognan BJ Pharm 2007)
Technologically based approaches to predict chemogenomic data (target selectivity for various ligands and ligand selectivity for various targets) will span pure ligand-based approaches (comparison of known ligands to predict their most probable targets), pure target-based approaches (comparison of targets or ligand-binding sites to predict their most likely ligands) or ultimately target-ligand based approaches (using experimental and predicted binding affinity matrices).
Basic questions :
1. Effectively we are talking about 'designing' molecules. (perhaps we should include ‘identifying’ –see below) This may not mean that more selective ligands are going to be ‘designed’, but simply that the observed selectivity profile of the compound will be compatible with a therapeutical usage. In addition, novel genomic targets could be better addressed after locating them in the target space and exploiting the associated chemical information. (Chemogenomic approaches to rational drug design by D Rognan)
2. Which families of compounds have been found to bind well with which families of protein targets?
3. Which families, and which members of which families of compounds have been found to have good ADME/Tox properties? Efficient cross-sectional comparison of results obtained via chemogenomic sequencing must be linked with test-subject specific genotypic metabolic profiles to enable rapid correlation of data to determine gene-specific metabolic toxicity profile data. This will provide phase I with selective, subject-specific, prospective metabolic profiling to enable industry to better circumvent potential toxicity/adverse events.
4. Which compounds have been found to have off-target effects.
5. Which families, and which members of which families of compounds have been found to be readily synthesizable? (sources: chembl.blogspot.com, www.kubinyi.de , Caron et al., 2001, Chemogenomic approaches to rational drug design by D Rognan )
Integrative Informatics Use Case
Integrative informatics involves bringing together data from discovery, development and the clinic to better understand the mechanism of action of a particular drug, or to identify biomakers as indicator of normal biologic or pathogenic process or pharmacological responses to a therapeutic intervention. Stratification markers are of particular interest, as they indicate which patients are likely to respond to drugs, which ones may have no response, and which patients are most likely to have adverse events.
The Alzheimer's Disease Neuroimaging Initiative (ADNI) has been undertaking a longitudinal natural history study. ADNI was designed to standardize AD clinical trial methodology involving imaging and biofluid biomarkers at over 57 sites in the US and Canada. The long-term goal is to qualify methods for early detection and disease progression. 200 normal patients have been enrolled for the last 3 years, 400 amnestic MCI have been enrolled for 3 years, and 200 Alzheimer's Disease patients have been enrolled for the last 2 years. Patients visit have samples taken every 6 months including clinical, blood and LP-CSF, 1.5T MRI, and cognitive tests. Some also have 3.0T MRI, FDG-PET, PiB-PET, and omics data analysis. The image data has already been analyzed using a variety of computational approaches.
This use proposes using the Translational Medicine Ontology to integrate the publicly available ADNI data, and attempt to stratify patients according to those whose condition is likely to deteriorate.
Animal Models Use Case
Animal models are useful in understanding human disease, treatments, and potential side effects. For example, behavioral conditions may be simulated or induced in laboratory animals to develop drugs for the treatment of these conditions in humans. Animal models to simulate a particular human disease may be spontaneous (naturally occurring) or induced by biological, chemical, or physical methods. (e.g. genetic modifications, behavior altering substances, cell implantations)
1. Which animal species has the closest genome for the pathway/target of this disease?
- search for gene/pathway implicated in disease in OMIM and Gene Ontology
- compare gene/pathway for various animal species to human gene/pathway with GMODs, Entrez Gene database, and OMIA
- compare past in-vivo experiments (internal data represented by the experiment ontology) with data from clinicaltrial.org
2. Do the in-vitro results support the in-vivo findings?
- compare past in-vitro experiments (internal data represented by the experiment ontology) with in-vivo experiments
3. Which genetic modifications would best simulate the human disease condition?
- Use GMODs?
4. Is PCP the best substance to simulate psychosis in neuroscience studies? What other substances may produce an effective model?
- compare past in-vivo experiments (internal data represented by the experiment ontology) with data from clinicaltrial.org
5. What safety issues have been seen in this target? (literature-based)
6. Are there predictive biomarkers for on-target and toxicological effects?
NOTE: Would be good to look at mice model to investigate stroke conditions
- http://www.helmholtz.de/en/research/health/the_latest_insights/mouse_as_a_model_for_stroke_patients/ - http://www.ncbi.nlm.nih.gov/pmc/articles/PMC484977/
-- Potential Sources of Information: --
LAMDHI – Link Animal Models to Human Disease http://www.lamhdi.org/about uses OMIM to link human genes
NINDS/Parkinson's http://www.ninds.nih.gov/research/parkinsonsweb/amr/animal_model.htm
NCBI - dbSNP and dbVAR
SNPedia Wiki - http://www.snpedia.com/index.php/SNPedia:FAQ
OpenTox - European Commission Seventh Framework Programme (FP7) http://opentox.org/
Predictive Tox Presentation: http://www.opentox.org/home/documents/presentations/presopentoxebinov2010/view
Look at IU Research: Bin Chen w/ Advisor David Wild
Home Page: http://chem2bio2rdf.wikispaces.com/
Datasets they are using: http://chem2bio2rdf.wikispaces.com/Datasets
Queries: http://chem2bio2rdf.wikispaces.com/multiple+sources
Publication Page: http://cheminfov.informatics.indiana.edu:8080/
Pharmacogenomics Use Case
Pharmacogenomics is the branch of pharmacology which deals with the influence of genetic variation on drug response in patients by correlating gene expression or single-nucleotide polymorphisms with a drug's efficacy or toxicity. By doing so, pharmacogenomics aims to develop rational means to optimise drug therapy, with respect to the patients' genotype, to ensure maximum efficacy with minimal adverse effects. Such approaches promise the advent of "personalized medicine"; in which drugs and drug combinations are optimized for each individual's unique genetic makeup. "(Guidance for Industry Pharmacogenomic Data Submissions" (PDF). U.S. Food and Drug Administration. March 2005)
Several clinical trials already capture cohort genetics. Crucial to the success of personalized medicine will be the development of decision support tools that can not only help capture relevant information, but also answer questions and make predictions about optimal therapeutic treatments given an individual’s background. The primary resource for PGx(pharmacogenetic, pharmacogenomic) knowledge is PharmGKB (http://www.pharmgkb.org/)
To develop, implement, and disseminate a public genotype-phenotype resource focused on pharmacogenetics and pharmacogenomics. In the short term, this resource will facilitate basic research. In the long term, it will impact how medicine is delivered. This resource will serve a broad community including geneticists, molecular biologists, pharmacologists, physicians, policy makers and the lay public. (missionof PharmGKB)
1. For what diseases is there evidence of a PGx response?
2. What drugs have known PGx associations? (almost all drugs have PGx associations) May want to restructure question to say, “what are the pharmacogenomic associations of the pharmacologic agents in question?
3. Which drugs may have a poor or negative outcome for certain genetic backgrounds?
4. What are the benefits of an administered pharmaceutical agent in the subject given their genetic variability? ie efficacy (Does the preparation even work in some people given their genotype ie. Metabolized too quickly by subject, or subject lacks various protein(s)/ target(s) necessary for physiologic response)
5. Which classes/groups of drugs are most effective ( provide the most desirable phenotype) (ie. pharmacodynamics, minimal side-effects, or maximal beneficial effect(s) on the disease indicating their use) for a given genotype? (ie. African American/White differences in efficacy of agents used as antihypertensives by JNC VII guidelines for the treatment of hypertension- genetic variability of subjects influences the agents indicated for use by clinicians, and influences the threshold for initiation of therapy)
6. What are the side effects of a given drug in a given genotype?
7. How can pharmacogenomic data be used to reduce toxicity in drug design? (Partly mentioned above) Efficient cross-sectional comparison of results obtained via chemogenomic sequencing must be linked with test-subject specific genotypic metabolic profiles, (pharmacogenomic data) to enable rapid correlation of data to determine gene-specific metabolic toxicity profile data. This will provide phase I with selective, subject-specific, prospective metabolic profiling to enable industry to better circumvent potential toxicity/adverse events. Pharmacogenomic data will enable investigators to catalyze scientific discovery in both pharmacogenetics/pharmacogenomics and biomedical informatics.(PharmGKB)
8.What are the specific pharmacogenomic criteria for a pharmaceutical agent that make it personalized? (These can be divided into those individuals whose genotype either A) limits or B)promotes use of the agent in question – Examples A) Warfarin – CYP2C9/ VCORK-1 genotyping predicting toxicity and limiting the necessary dose in a portion of the population to avoid bleeding, and B)HER2 receptor positive breast cancer – showing the benefit only in those individuals who are receptor positive )