Biomedical Ontologies What are they (for) ?

1
Biomedical Ontologies What are they (for) ?

• Stefan Schulz
• Medical Informatics
• Research Group
• UniversityMedical Center
• Freiburg, Germany

2
Understanding / Semantic Interoperability
HealthCare
data
data
Enables understanding between human and
computational agents
PublicHealth
Consumers
data
data
data
BiomedicalResearch
Common language Ontologies and Terminology
Systems
3
Ontologies and Terminology Systems

• aka Knowledge Organization Systems Systems that
  support semantic interoperability by
  communicating and processing information
• In a structured form
• Well-defined
• Unambiguous
• Processable by machines
• Understandable by humans
• Life Sciences major focus for the development of
  ontologies and terminological systems

4
Literature on Biomedical Terminologies and
Ontologies
5
Purpose of this Talk
Formal

• What are Ontologies
• What are they for ?

6
Structure of this talk

• Introduction - Current Systems
• Terminological Clarification
• What do Formal Ontologies Represent ?
• Terminologies vs. Formal Ontologies
• Practice of Good Ontology
• Outlook

7
Structure of this talk

• Introduction - Current Systems
• Terminological Clarification
• What do Formal Ontologies Represent ?
• Terminologies vs. Formal Ontologies
• Practice of Good Ontology
• Outlook

8
A cruise through the archipelago of systems for
biomedical knowledge organization
FBcv
9
MeSH Medical Subject Headings
MeSHMedical Subject Headings
10
(No Transcript)
11
Hierarchical principle broader term / narrower
term (not a taxonomy)
12
(No Transcript)
13
(No Transcript)
14
MeSH Medical Subject Headings
GOGene Ontology
15
(No Transcript)
16
(No Transcript)
17
Part of (partonomy)
Is a (taxonomy)
18
MeSH Medical Subject Headings
ICDInternational Classification of Diseases
19
(No Transcript)
20
(No Transcript)
21
Class / subclass Relation (is_a)
22
MeSH Medical Subject Headings
SNOMED Clinical Terms
23
(No Transcript)
24
SNOMED CT Facts (I)

• SNOMED CT is a terminology
• consisting of terms used in health health care,
• attached to concept codes with multiple terms per
  code
• structured according to logic-based
  representation of meanings
• increasingly guided by ontological principles
• Current size
• 283,000 Concepts
• 732,000 Terms
• 923,000 Concept Concept Relations

25
SNOMED CT Facts (II)

• Since 2007 Maintained by IHTSDO (International
  Health Terminology standards development
  organization)
• Members Australia, Canada, Denmark, Lithuania,
  The Netherlands, New Zealand, Sweden, UK, USA.
• Annual budget 5 M

26
Different Purposes Heterogeneous Approaches
27
Different Purposes Heterogeneous Approaches

• MeSH Medical Subject Headings Hierarchy
  (broader / narrower) of descriptors, used for
  indexing biomedical publications for literature
  retrieval support
• GO Gene OntologyHierarchy (is_a / part_of) of
  controlled terms for describing gene an gene
  product properties
• ICD International Classification of
  DiseasesStrict Hierarchy of non-overlapping
  classes for classifying statistically relevant
  health conditions
• SNOMED CT Systematized Nomenclature of Medicine
  Clinical Terms Hierarchical system of
  concepts with (partially) logic-based concept
  definitions

28
Other Biomedical Knowledge Organization Systems
Medicine
Source UMLS
International Classification of Primary
Care International Classification of Primary Care
2nd Edition International Statistical
Classification of Diseases and Related Health
Problems JAMAS Japanese Medical Thesaurus
(JJMT) Library of Congress Subject Headings LOINC
2.15 Master Drug Data Base McMaster University
Epidemiology Terms Medical Dictionary for
Regulatory Activities Terminology
(MedDRA) Medical Entities Dictionary Medical
Subject Headings MEDLINE (1996-2000) MEDLINE
(2001-2006) MedlinePlus Health Topics_2004_08_14 M
icromedex DRUGDEX Multum MediSource Lexicon NANDA
nursing diagnoses definitions
classification National Drug Data File Plus
Source Vocabulary National Drug File - Reference
Terminology National Library of Medicine Medline
Data NCBI Taxonomy
AI/RHEUM Alcohol and Other Drug
Thesaurus Alternative Billing Concepts Beth
Israel Vocabulary Canonical Clinical Problem
Statement System Clinical Classifications
Software Clinical Terms Version 3 (CTV3) (Read
Codes) Common Terminology Criteria for Adverse
Events COSTAR COSTART CRISP Thesaurus Current
Dental Terminology 2005 (CDT-5) Current
Procedural Terminology Diseases
Database DSM-III-R DSM-IV DXplain Gene
Ontology HCPCS Version of Current Dental
Terminology 2005 (CDT-5) HCPCS Version of Current
Procedural Terminology (CPT) Healthcare Common
Procedure Coding System HL7 Vocabulary Version
2.5 HL7 Vocabulary Version 3.0 Home Health Care
Classification HUGO Gene Nomenclature ICD10 ICD-9-
CM ICPC ICPC2 - ICD10 Thesaurus ICPC2-ICD10
Thesaurus
NCI SEER ICD Neoplasm Code Mappings NCI
Thesaurus Neuronames Brain Hierarchy Nursing
Interventions Classification Nursing Outcomes
Classification Omaha System Online Congenital
Multiple Anomaly/Mental Retardation
Syndromes Online Mendelian Inheritance in
Man Patient Care Data Set Perioperative Nursing
Data Set Pharmacy Practice Activity
Classification Physician Data Query Physicians'
Current Procedural Terminology Quick Medical
Reference (QMR) Read thesaurus Read thesaurus
Americanized Synthesized Terms RXNORM
Project SNOMED-2 SNOMED Clinical Terms SNOMED
International Standard Product Nomenclature Thesau
rus of Psychological Index Terms The Universal
Medical Device Nomenclature System
(UMDNS) UltraSTAR UMLS Metathesaurus University
of Washington Digital Anatomist USP Model
Guidelines Veterans Health Administration
National Drug File WHO Adverse Reaction
Terminology WHOART
29
Other Biomedical Knowledge Organization Systems
Biology (OBO)
30
Structure of this talk

• Introduction - Current Systems
• Terminological Clarification
• What do Formal Ontologies Represent ?
• Terminologies vs. Formal Ontologies
• Practice of Good Ontology
• Outlook

31
Unresolved Terminological Confusion

• Knowledge Organization Systems artifacts for
  ordering domain entities, relating word meanings
  or providing semantic reference

• Vocabularies
• Terminologies
• Thesauri
• Concept Systems
• Classifications
• (Formal) Ontologies

32
Unresolved Terminological Confusion

• Different scientific traditions Biology,
  Medicine, Philosophy, Logic, Linguistics,
  Library and Information Science, Computer
  Science, Cognitive Science, International
  Terminology norms
• Different philosophical schools of thinking
  Platonism, Aristotelian Realism, Conceptualism,
  Relativism, Idealism, Postmodernism,
  Constructivism, Nominalism, Tropism,

33
Components of Knowledge Organization Systems
Hierarchically ordered Nodes and Links
Formal or informal Definitions
Dictionaries of Natural language Terms
domain or region of DNA GENIA A substructure
of DNA molecule which is supposed to have a
particular function, such as a gene, e.g., c-jun
gene, promoter region, Sp1 site, CA repeat. This
class also includes a base sequence that has a
particular function.

• Benign neoplasm of heart
• Benign tumor of heart
• Benign tumour of heart
• Benign cardiac neoplasm
• Gutartiger Herzumor
• Gutartige Neubildung am Herzen
• Gutartige Neubildung Herz
• Gutartige Neoplasie des Herzens
• Tumeur bénigne cardiaque
• Tumeur bénigne du cœur
• Neoplasia cardíaca benigna
• Neoplasia benigna do coração
• Neoplasia benigna del corazón
• Tumor benigno do corazón

Peptides MeSH Members of the class of
compounds composed of AMINO ACIDS joined together
by peptide bonds between adjacent amino acids
into linear, branched or cyclical structures.
OLIGOPEPTIDES are composed of approximately 2-12
amino acids. Polypeptides are composed of
approximately 13 or more amino acids. PROTEINS
are linear polypeptides that are normally
synthesized on RIBOSOMES.
19429009chronic ulcer of skin116680003is
a64572001disease 116676008associated
morphology 405719001chronic ulcer
363698007finding site 39937001skin
structure
34
What do the nodes in Formal Ontologies /
Terminological Systems stand for?
universals
names
categories
types
sets
descriptors
synsets
sorts
entities
properties
classes
terms
descriptors
concepts
35
Ontology Gradient or crisp boundary ?
Terminology
Ontology

Information Model
36
Ontology Gradient or crisp boundary ?
Terminology
Formal Ontology
Information Model
37
Organizing the world
Terminology
Formal Ontology

• Set of terms representing the system of concepts
  of a particular subject field.  (ISO 1087)

Ontology is the study of what there is. Formal
ontologies are theories that attempt to give
precise mathematical formulations of the
properties and relations of certain entities.
(Stanford Encyclopedia of Philosophy)
38
Structure of this talk

• Introduction - Current Systems
• Terminological Clarification
• What do Formal Ontologies Represent ?
• Terminologies vs. Formal Ontologies
• Practice of Good Ontology
• Outlook

39
Terminologies start with human language
Terminology
Formal Ontology

• Set of terms representing the system of concepts
  of a particular subject field.  (ISO 1087)

Ontology is the study of what there is. Formal
ontologies are theories that attempt to give
precise mathematical formulations of the
properties and relations of certain entities.
(Stanford Encyclopedia of Philosophy)
40
Semantic Reference
Entities of Language (Terms)
Shared / Meanings / Entities of Thought
(Concepts)
benign neoplasm of heart gutartige
Neubildung des Herzmuskels neoplasia cardíaca
benigna
41
Example UMLS (mrconso table)
Shared / Meanings / Entities of Thought
Entities of Language (Terms)

• C0153957ENGPL0180790PFS1084242YA1141630
  MTHPNU001287benign neoplasm of heart0N
• C0153957ENGPL0180790VCS0245316NA0270815
  ICD9CMPT 212.7Benign neoplasm of heart0N
• C0153957ENGPL0180790VCS0245316NA0270817
  RCDSYB727. Benign neoplasm of heart3N
• C0153957ENGPL0180790VOS1446737YA1406658
  SNMIPT D3-F0100Benign neoplasm of heart,
  NOS3N
• C0153957ENGSL0524277PFS0599118NA0654589
  RCDAEPTB727.Benign tumor of heart3N
• C0153957ENGSL0524277VOS0599510NA0654975
  RCDPTB727. Benign tumour of heart3N
• C0153957ENGSL0018787PFS0047194YA0066366
  ICD10PSD15.1Heart3Y
• C0153957ENGSL0018787VOS0900815YA0957792
  MTHMMU003158Heart lt3gt0Y
• C0153957ENGSL1371329PFS1624801NA15830561
  0004245MDRLT10004245Benign cardiac
  neoplasm3N
• C0153957GERPL1258174PFS1500120YA1450314
  DMDICD10PT D15.1Gutartige Neubildung
  Herz1N
• C0153957SPAPL2354284PFS2790139NA2809706
  MDRSPALT 10004245Neoplasia cardiaca
  benigna3N

Unified Medical Language System, Bethesda, MD
National Library of Medicine, 2007
http//umlsinfo.nlm.nih.gov/
42
Example UMLS (mrrel table)

• C0153957A0066366AUIPARC0348423A0876682AUI
  R06101405ICD10ICD10N
  
• C0153957A0066366AUIRQ C0153957A0270815AUI
  default_mapped_ fromR03575929NCISEERNCISEER
  N
• C0153957A0066366AUISY C0153957A0270815AUI
  uniquely_mapped_ to R03581228NCISEERNCISEER
  N
• C0153957A0270815AUIRQ C0810249A1739601AUI
  classifies R00860638CCSCCSN
• C0153957A0270815AUISIBC0347243A0654158AUI
  R06390094
  ICD9CMICD9CMNN
• C0153957A0270815CODERNC0685118A3807697SCUI
  mapped_to R15864842SNOMEDCTSNOMEDC
  TYN
• C0153957A1406658AUIRL C0153957A0270815AUI
  mapped_from R04145423SNMISNMIN
• C0153957A1406658AUIRO C0018787A0357988AUI
  location_of R04309461SNMISNMIN
• C0153957A2891769SCUICHDC0151241A2890143SCUI
  isa R1984122047189027SNOMEDCTS
  NOMEDCT0YN

43
Example UMLS
Shared / Meanings / Entities of Thought
Shared / Meanings / Entities of Thought

• C0153957A0066366AUIPARC0348423A0876682AUI
  R06101405ICD10ICD10N
  
• C0153957A0066366AUIRQ C0153957A0270815AUI
  default_mapped_ fromR03575929NCISEERNCISEER
  N
• C0153957A0066366AUISY C0153957A0270815AUI
  uniquely_mapped_ to R03581228NCISEERNCISEER
  N
• C0153957A0270815AUIRQ C0810249A1739601AUI
  classifies R00860638CCSCCSN
• C0153957A0270815AUISIBC0347243A0654158AUI
  R06390094
  ICD9CMICD9CMNN
• C0153957A0270815CODERNC0685118A3807697SCUI
  mapped_to R15864842SNOMEDCTSNOMEDC
  TYN
• C0153957A1406658AUIRL C0153957A0270815AUI
  mapped_from R04145423SNMISNMIN
• C0153957A1406658AUIRO C0018787A0357988AUI
  location_of R04309461SNMISNMIN
• C0153957A2891769SCUICHDC0151241A2890143SCUI
  isa R1984122047189027SNOMEDCTS
  NOMEDCT0YN

Semantic relations
INFORMAL
44
Formal Ontology represents the world
Terminology
Formal Ontology

• Set of terms representing the system of concepts
  of a particular subject field.  (ISO 1087)

Ontology is the study of what there is (Quine).
Formal ontologies are theories that attempt to
give precise mathematical formulations of the
properties and relations of certain entities.
(Stanford Encyclopedia of Philosophy)
45
Organizing Entities
46
Organizing Entities
Entity Types
The type benign neoplasm of heart
abstract
Universals, classes, (Concepts)
The benign neoplasm of my heart
Entities of the World
concrete
Particulars, instances
47
Organizing Entities
Entity Types
The type benign neoplasm of heart
abstract
Universals, classes, (Concepts)
Entities of Language
Terms, names
The benign neoplasm of my heart
Entities of the World
concrete
The string benign neoplasm of heart
Particulars, instances
48
Organizing Entities
(the complication of my) benign heart tumor
(die Komplikation meines) Gutartigen
Herztumors
represents
49
Organizing Entities
represents
(the) benign heart tumor (is congenital)
(die Komplikation meines) Gutartigen
Herztumors
Terms, names
50

Entities of Language
are stored in dictionaries and represented by
terminologies
51

Database systems / information models store
references to
Entities of the World
52
Entity Types
are organized in formal ontologies
53
Hierarchical framework for entity types

• Taxonomy relates types and subtypes
• Tumor of Heart is_a Tumor equivalent to
• All instances of Tumor of Heart are instances of
  Tumor(without exceptions)
• Relations
• instance_of relates instances with types, all
  others relate instances (e.g. part_of) or are
  derived from them (e.g. is_a)
• Definitions describe what is always true for all
  instances of a type
• Tumor of Heart has_location Heart All
  instances of Tumor of Heart are located in some
  Heart

54
Type / Subtype Hierarchy
Tumor of Heart
Benign Tumor
Is_a
Is_a
Is_a
Benign Tumor of Heart
Malignant Tumor of Heart
55
A classification view on Formal Ontologies
World
56
Hierarchies, Types, Classes, Individuals
World
57
Hierarchies, Types, Classes, Individuals
World
58
Hierarchies, Types, Classes, Individuals
Type 1
World
59
Hierarchies, Types, Classes, Individuals
Formal Ontology
Type 1
Is_a
Is_a
Is_a
Subtype 1.2
Subtype 1.1
Subtype 1.3
World
60
Hierarchies, Types, Classes, Individuals
Formal Ontology
InflammatoryDisease
61
Hierarchies, Types, Classes, Individuals
Formal Ontology
InflammatoryDisease
Is_a
Is_a
Is_a
Hepatitis
Gastritis
Pacreatitis
62
Hierarchies, Types, Classes, Individuals
Formal Ontology
InflammatoryDisease
Is_a
Is_a
Is_a
Hepatitis
Gastritis
Pacreatitis
63
Hierarchies, Types, Classes, Individuals
Formal Ontology
InflammatoryDisease
Is_a
Is_a
Is_a
Hepatitis
Gastritis
Pacreatitis
64
Relations and Definitions
Formal Ontology
InflammatoryDisease
hasLocation
Is_a
Liver
Hepatitis
65
Relations and Definitions
Formal Ontology
InflammatoryDisease
hasLocation
Is_a
Liver
Hepatitis
66
Relations and Definitions
Formal Ontology
InflammatoryDisease
hasLocation
Is_a
Population
Liver
Hepatitis
causedby
Is_a
Population of Virus
Viral Hepatitis
67
Languages for formal ontologies

• Natural Language
• Logic

Every hepatitis is an inflammatory disease that
is located in some liver Every inflammatory
disease that is located in some liver is an
hepatitis
?x instanceOf(x, Hepatitis) ? instanceOf(x,
Inflammation) ? ?y
instanceOf(y, Liver) ? hasLocation(x,y)
Logic is computable it supports machine
inferences but
it only scales up if it has a very limited
expressivity
68
Structure of this talk

• Introduction - Current Systems
• Terminological Clarification
• What do Formal Ontologies Represent ?
• Terminologies vs. Formal Ontologies
• Practice of Good Ontology
• Outlook

69
Terminologies vs. Formal Ontologies
Terminologies Formal Ontologies

• Describe entities of reality as they generically
  are independent of human language
• Types represent the generic properties of
  world entities
• Relations rigid, exactly defined, quantified
  relationships between particulars
• Description patternfor all instance of Type1
  there is some

• Describe Meaning of human language units
• Concepts aggregate (quasi)-synonymous terms
• Relations informal, elastic Associations between
  Concepts ..
• Description patternConcept1 Relation Concept2

70
Example Hepatitis - Liver
Terminologies Formal Ontologies

• Type Hepatitis
• Description
• Every hepatitis is an inflammatory disease that
  is located in some liver Every inflammatory
  disease that is located in some liver is an
  hepatitis

• Concept Hepatitis Hepatitis (D),
  Leberentzündung (D), hepatitis (E), hépatite (F)
• Concept Liver Leber (D), liver (E), foie (F)
• Relations
• Hepatitis hasLocation Liver
• Hepatitis isA - Inflammation

71
Example Hand - Thumb
Terminologies Formal Ontologies

• Type Thumb
• Description

• Concept Hand Hand (D), hand (E), main (F)
• Concept Thumb Daumen (D), thumb (E), pouce
  (F)
• Relations
• Hand hasPart Thumb
• Thumb partOf Hand

Every thumb is part of some hand Every hand
has some thumb as part
?
72
Example Aspirin - Headache
Terminologies Formal Ontologies

• Type Aspirin
• Description

• Concept Aspirin Aspirin (D,E),
  Acetylsalicylsäure (D), ASS (D), acetylsalicylic
  acid (E), Acide acétylsalicylique(F)
• Concept Headache Kopfschmerz (D), headache
  (E), céphalée(F)
• Relation
• Aspirin treats Headache

For every portion of aspirin there is some
disposition for treating headache
73
Strengths of Formal Ontologies

• Exact, logic-based descriptions of entity types
  that are instantiated by real-world objects,
  processes, states
• Representation of stable, context-independent
  accounts of reality
• Use of formal reasoning methods using tools and
  approaches from the AI / Semantic Web community

74
Deficit of Nomenclatures / Terminologies

• D5-46210 Acute appendicitis, NOS
• D5-46100 Appendicitis, NOS
• G-A231 Acute
• M-41000 Acute inflammation, NOS
• G-C006 In
• T-59200 Appendix, NOS
• G-A231 Acute
• M-40000 Inflammation
• G-C006 In
• T-59200 Appendix, NOS

SNOMED INTERNATIONAL
75
Formal-ontological descriptions Advantages

• Different description of the same thing can be
  automatically mapped to a canonic description by
  a logic-based reasoning device
• Meaning of defined classes can be unambiguously
  expressed

76
Formal Ontologies Limitations (I)

• Only suitable to represent shared,
  uncontroversial meaning of a domain vocabulary
• Supports universal statements about instances of
  a type
• All Xs are Ys
• For all Xs there is some Y
• Properties of types are properties of all
  entities that instantiate these types (strict
  inheritance)

77
Formal Ontologies Limitations (II)

• Representation of context dependent knowledge
• Allergic Rhinitis is a common disorder (in
  Europe)
• Representation of probabilistic knowledge
• 95 of people infected with viral hepatitis
  recover
• Smoking is a cardiovascular risk factor
• Default / canonic knowledge
• Adult humans have 32 teeth
• Dispositions
• Oxazepam is indicated for anxiety disorders
• Aspirin affects the gastric mucosa

Ontology ? Knowledge Representation
78
Continuum of knowledge
Universally accepted assertions
Consolidated but context-dependent facts
Hypotheses, beliefs, statistical associations
Domain Knowledge
79
Formal Ontology !
Universally accepted assertions
Consolidated but context-dependent facts
Hypotheses, beliefs, statistical associations
Domain Knowledge
80
Structure of this talk

• Introduction - Current Systems
• Terminological Clarification
• What do Formal Ontologies Represent ?
• Terminologies vs. Formal Ontologies
• Practice of Good Ontology
• Outlook

81
Practice of Good Ontology

• Learning good ontology practice from bad
  ontologies

82
Dont mix up universals (Concepts, Classes) with
individuals (Instances)
Is_a subclass_ofTaxonomic Subsumption

• subclass-of (Motor Neuron, Neuron) (FMA,
  OpenGALEN)
• Is_a (Motor Neuron, Neuron)
• instance-of (Motor Neuron, Neuron) (FlyBase)
• But
• instance-of (my Hand, Hand)
• instance-of (this amount of insulin, Insulin)
• instance-of (Germany, Country)
• not instance of (Heart, Organ)
• not instance of (Insulin, Protein)

Instance_of Class Membership
83
Keep in mind the meaning of taxonomic subsumption
(Is_a)

• Is_a (A, B) means
• All instances of A are instances of B
• Is_a (Geographic Area, Spatial Concept)
• Is_a (Body Structure, SNOMED concept) (UMLS-SN)
• ? instance-of (my heart, SNOMED concept)
  (SNOMED CT)
• Is_a (Protein Family or Group, Protein) (GENIA)
• Is_a (Absence of liver or gallbladder NOS
  Congenital absence of liver and gallbladder)
  (SNOMED CT)

No subclassing without inheritance!
84
Dont use superclasses to express roles

• Is_a (Fish, Animal)
• Is_a (Fish, Food) ??
• Is_a (Acetylsalicylic Acid, Salicylate)
• Is_a (Acetylsalicylic Acid, Analgetic Drug) ??

Be aware of the rigidity of entity types
85
Partition the ontology by principled upper level
categories
Example DOLCEs Upper Ontology
Endurant (Continuant) Physical Amount of
matter Physical object Feature Non-Physical
Mental object Social object Perdurant
(Occurrent) Static State Process Dynamic A
chievement Accomplishment
Quality Physical Qualities Spatial
location Temporal Qualities Temporal
location Abstract Qualities Abstract Qu
ality region Time region Space region Color
region
Source S. Borgo ISTC-CNR
86
Limit to a parsimonious set of semantically
precise Basic Relations
Barry Smith, Werner Ceusters, Bert Klagges, Jacob
Köhler, Anand Kumar, Jane Lomax, Chris Mungall,
Fabian Neuhaus, Alan L Rector and Cornelius
Rosse. Relations in biomedical ontologies. Genome
Biology, 6(5), 2005.
87
Avoid idiosyncratic categorization
Physical object (8) Device Domestic, office
and garden artefact Fastening () Procedur
e (23) Administrative procedure Community
health procedure () Qualifier value
(52) Action Additional dosage
instructions () Record artifact Record
organizer Record type Situation with explicit
context (17) A/N risk factors Critical
incident factors () Social context
(10) Community Family Group
() Special concept Namespace
concept Navigational concept Non-current
concept Specimen (45) Biopsy sample Body
substance sample Cardiovascular
sample () Staging and scales (6) Assessment
scales Endometriosis classification of
American Fertility Society () Substance
(11) Allergen class Biological
substance Body substance ()
Body structure (10) Acquired body
structure Anatomical organizational
pattern () Clinical finding
(22) Administrative statuses Adverse incident
outcome categories () Environment or
geographical location Environment Geogr.
and/or political region of the world Event
(19) Abuse Accidental event Bioterrorism
related event () Linkage concept Attribute
Link assertion Observable entity Age AND/OR
growth period Body product observable () Cli
n. history / examination observable (21) Device
observable Drug therapy observable Feature of
Entity () Organism (11) Animal Chromista
Infectious agent () Pharmaceutical /
biologic product (58) Alcohol
products Alopecia preparation Alternative
medicines () Physical force
(21) Altitude Electricity ()
88
The Celestial Emporium of Benevolent Knowledge

1. stray dogs
2. those that are included in this classification
3. those that tremble as if they were mad
4. innumerable ones
5. those drawn with a very fine camel's hair brush
6. others
7. those that have just broken a flower vase
8. those that resemble flies from a distance"

• Jorge Luis Borges
• "On those remote pages it is written that
  animals are divided into
• those that belong to the Emperor
• embalmed ones
• those that are trained
• suckling pigs
• mermaids
• fabulous ones

89
Be aware of ambiguities

• Institution may refer to
• (abstract) institutional rules
• (concrete) things instituted
• act of instituting sth.
• Tumor
• evolution of a tumor as a disease process
• having a tumor as a pathological state
• tumor as a physical object
• Gene
• a (physical) sequence of nucleotides on a DNA
  chain
• a collection of (1)
• A piece of information conveyed by (1)

90
Dont mix up ontology with epistemiology

• Is_a (Infection of unknown origin, Infection)
• Is_a (Newly diagnosed diabetes, Diabetes)
• Is_a (Family history of diabetes, Diabetes)

what is
what sth. knows about
91
Dont mix up Ontology IDs with Terms

• Glycerin Kinase
• Glycerokinase
• GK
• Glyzerinkinase

92
what is
how it is expressed in human language
what sth. knows about
93
Dont underestimate Ontology Maintenance

• Formal Ontologies must always be maintained
• consistent (free of logic contradiction)
  prerequisite for machine reasoning
• adequate (correctly describe the domain)
  prerequisite to prevent erroneous deductions
• Maintenance load is much higher than with
  terminologies.
• Ontology maintenance is mainly task of domain
  experts. IT staff has supportive function
• Typical design and maintenance errors

94
Structure of this talk

• Introduction - Current Systems
• Terminological Clarification
• What do Formal Ontologies Represent ?
• Terminologies vs. Formal Ontologies
• Practice of Good Ontology
• Outlook

95
Outlook

• Ontology often used a buzzword for nontologies
  but
• Formal ontological principles increasingly govern
  the construction of Life Science Knowledge
  Organization Systems
• Users / domain expert must be heavily involved
  into ontology engineering and maintenance
• Insufficient evidence
• Which use cases require formal ontologies
• In which cases informal terminology systems are
  sufficient?
• Which cases require both ?
• Can existing terminologies be ontologized?
• Can terminologies and ontologies co-exist ?
• The outcome of the existing legacy systems move
  toward principled ontologies is still open

96
SNOMED CT

• One huge system
• Impressive domain coverage
• Considerable investments, important stakeholders
• Increasing number of (clinical) users
• Other use cases mainly unexplored (basic
  research, clinical trials)
• Increasing mappings to existing terminologies
• Legacy hybrid of terminology, ontology,
  information model
• Overall structure idiosyncratic, disorganized
• Some major architectural weaknesses
• Unreflected use of logic, unintended entailments
• Major redesign necessary formal foundations,
  editing guidelines, quality control procedures
• Risk uncontrollable proliferation, loss of
  expressiveness,
• Chances Positive input by user groups

97
Open Biomedical Ontologies (OBO)

• Many focused ontologies
• Increasing number of annotated sources
• Broad range organisms anatomies (plants,
  animals) pathways biomedical investigations
  cells development protein sequence
• Convergence to standardized syntax and semantics
• Increasingly using formal ontology principles
• Public

98
Thank you!Contactstschulz_at_uni-freiburg.de

• Stefan Schulz
• Medical Informatics
• Research Group
• UniversityMedical Center
• Freiburg, Germany