1. Ontologies for Data Integration: A Semantic Web Perspective Training Course Olivier Bodenreider Lister Hill National Center for Biomedical Communications Bethesda, Maryland - USA May 24, 2006 Schloß Dagstuhl Biomedical Ontology in

2. 2 Outline u u From terminology integration to information integration Unified Medical Language System (UMLS) u u UMLS in use: Mapping across terminologies u u Beyond UMLS: Towards a Biomedical Knowledge Repository

3. From terminology integration to information integration Unified Medical Language System (UMLS)

4. 4 What does UMLS stand for? u Unified u Medical u Language u System UMLS ® Unified Medical Language System ® UMLS Metathesaurus ®

5. 5 Motivation u Started in 1986 u National Library of Medicine u “Long-term R&D project” «[…] the UMLS project is an effort to overcome two significant barriers to effective retrieval of machine-readable information. The first is the variety of ways the same concepts are expressed in different machine-readable sources and by different people. The second is the distribution of useful information among many disparate databases and systems.»

6. Overview through an example

7. 7 Addison’s disease u Addison's disease is a rare endocrine disorder u Addison's disease occurs when the adrenal glands do not produce enough of the hormone cortisol u For this reason, the disease is sometimes called chronic adrenal insufficiency, or hypocortisolism

8. 8 Adrenal insufficiency Clinical variants u Primary / Secondary l Primary: lesion of the adrenal glands themselves l Secondary: inadequate secretion of ACTH by the pituitary gland u Acute / Chronic u Isolated / Polyendocrine deficiency syndrome ACTH

9. 9 Addison’s disease: Symptoms u Fatigue u Weakness u Low blood pressure u Pigmentation of the skin (exposed and non- exposed parts of the body) u …

10. 10 AD in medical vocabularies u Synonyms: different terms l Addisonian syndrome l Bronzed disease l Addison melanoderma l Asthenia pigmentosa l Primary adrenal deficiency l Primary adrenal insufficiency l Primary adrenocortical insufficiency l Chronic adrenocortical insufficiency u Contexts: different hierarchies symptoms clinical variants eponym

11. 11 Organize terms u Synonymous terms clustered into a concept u Preferred term u Unique identifier (CUI) Addison's disease Addison DiseaseMeSHD000224 Primary hypoadrenalismMedDRA10036696 Primary adrenocortical insufficiencyICD-10E27.1 Addison's disease (disorder)SNOMED CT363732003 C0001403

12. Diseases of the endocrine system Diseases of the Adrenal Glands Addison’s Disease Diseases/Diagnoses SNOMED International

13. Endocrine Diseases Adrenal Gland Diseases Addison’s Disease Diseases MeSH Adrenal Gland Hypofunction

14. Endocrine disorder Adrenal disorder Adrenal cortical disorder Adrenal cortical hypofunction Addison’s Disease AOD

15. Endocrine disorder Disorder of adrenal gland Hypoadrenalism Adrenal Hypofunction Corticoadrenal insufficiency Addison’s Disease Read Codes

16. Primary adrenocortical insufficiency Other disorders of adrenal gland Disorders of other endocrine gland ICD-10

17. 17 Organize concepts u Inter-concept relationships: hierarchies from the source vocabularies u Redundancy: multiple paths u One graph instead of multiple trees (multiple inheritance) A BDEHDE B GH EFH C BC A EFD GH

18. Adrenal Cortex Diseases Hypoadrenalism Adrenal Gland Hypofunction Adrenal cortical hypofunction Endocrine Diseases Adrenal Gland Diseases organize concepts Addison’s Disease UMLS SNOMED MeSH AOD Read Codes

19. 19 Relate to other concepts u Additional hierarchical relationships l link to other trees l make relationships explicit u Non-hierarchical relationships u Co-occurring concepts u Mapping relationships

20. Endocrine Diseases Adrenal Gland Diseases Adrenal Cortex Diseases Hypoadrenalism Adrenal Gland Hypofunction Adrenal cortical hypofunction Addison’s Disease Adrenal Cortex Dysfunction Adrenal Dysfunction Addison’s disease due to autoimmunity Secondary hypocortisolism Other disorders of adrenal gland Disorders of other endocrine gland Adrenal Glands Adrenal Cortex Endocrine System Endocrine Glands Abdominal organ Diseases relate to other concepts

21. 21 Categorize concepts u High-level categories (semantic types) u Assigned by the Metathesaurus editors u Independently of the hierarchies in which these concepts are located Disease or Syndrome Endocrine Diseases Adrenal Gland Diseases Addison’s Disease Diseases Adrenal Gland Hypofunction

22. 22 How do they do that? u Lexical knowledge u Semantic pre-processing u UMLS editors

23. 23 Lexical knowledge Adrenal gland diseases Adrenal disorder Disorder of adrenal gland Diseases of the adrenal glands C0001621

24. 24 Semantic pre-processing u Metadata in the source vocabularies u Tentative categorization u Positive (or negative) evidence for tentative synonymy relations based on lexical features

25. 25 Additional knowledge: UMLS editors Adrenal Gland Diseases Adrenal Cortex Diseases Adrenal Cortex Dysfunction Hypoadrenalism Adrenal Gland Hypofunction Adrenal cortical hypofunction Addison’s Disease Other disorders of adrenal gland

26. 26 UMLS: 3 components u SPECIALIST Lexicon l 200,000 lexical items l Part of speech and variant information u Metathesaurus l 5M names from over 100 terminologies l 1M concepts l 16M relations u Semantic Network l 135 high-level categories l 7000 relations among them Lexical resources Ontological resources Terminological resources

27. UMLS Metathesaurus

28. 28 Source Vocabularies u 139 source vocabularies l 17 languages u Broad coverage of biomedicine l 5.1M names l 1.3M concepts l 16M relations u Common presentation (2006AB)

29. 29 Addison’s Disease: Concept Addison’s Disease C0001403 ADRENAL INSUFFICIENCY (ADDISON'S DISEASE) ADRENOCORTICAL INSUFFICIENCY, PRIMARY FAILURE Addison melanoderma Melasma addisonii Primary adrenal deficiency Asthenia pigmentosa Bronzed disease Insufficiency, adrenal primary Primary adrenocortical insufficiency Addison's, disease MALADIE D'ADDISON - French Addison-Krankheit - German Morbo di Addison - Italian DOENCA DE ADDISON - Portuguese ADDISONOVA BOLEZN' - Russian ENFERMEDAD DE ADDISON - Spanish A disease characterized by hypotension, weight loss, anorexia, weakness, and sometimes a bronze-like melanotic hyperpigmentation of the skin. It is due to tuberculosis- or autoimmune-induced disease (hypofunction) of the adrenal glands that results in deficiency of aldosterone and cortisol. In the absence of replacement therapy, it is usually fatal. SNOMED MeSH AOD Read Codes … Disease or Syndrome

30. 30 Metathesaurus Concepts u Concept(> 1.3M)CUI l Set of synonymous concept names u Term(> 4.6M)LUI l Set of normalized names u String(> 5.1M)SUI l Distinct concept name u Atom(> 6.2M)AUI l Concept name in a given source (2006AB) A0000001 headache (source 1) A0000002 headache (source 2) S0000001 A0000003 Headache (source 1) A0000004 Headache (source 2) S0000002 L0000001 A0000005 Cephalgia (source 1) S0000003 L0000002 C0000001

31. 31 Cluster of synonymous terms Concept C0001621 […] Term L0001621 S0011232 Adrenal Gland Diseases S0011231 Adrenal Gland Disease S0000441 Disease of adrenal gland S0481705 Disease of adrenal gland, NOS S0220090 Disease, adrenal gland S0044801 Gland Disease, Adrenal Term L0041793 S0860744 Disorder of adrenal gland, unspecified S0217833 Unspecified disorder of adrenal glands […] Term L0368399 S0586222 Adrenal disease S0466921 ADRENAL DISEASE, NOS […] Term L0181041 S0632950 Disorder of adrenal gland S0354509 Adrenal Gland Disorders […] Term L0161347 S0225481 ADRENAL DISORDER S0627685 DISORDER ADRENAL (NOS) […] Term L1279026 S1520972 Nebennierenkrankheiten GER S0226798 SURRENALE, MALADIES Term L0162317 FRE

32. 32 Metathesaurus Evolution over time u Concepts never die (in principle) l CUIs are permanent identifiers u What happens when they do die (in reality)? l Concepts can merge or split l Resulting in new concepts and deletions Addison's disease C0001403 Addison's disease, NOS C0271735 199219931994199519961997199819992004 …

33. 33 Metathesaurus Relationships u Symbolic relations:~9 M pairs of concepts u Statistical relations :~7 M pairs of concepts (co-occurring concepts) u Mapping relations:100,000 pairs of concepts u Categorization: Relationships between concepts and semantic types from the Semantic Network

34. 34 Symbolic relations u Relation l Pair of “atom” identifiers l Type l Attribute (if any) l List of sources (for type and attribute) u Semantics of the relationship: defined by its type [and attribute] Source transparency: the information is recorded at the “atom” level

35. 35 Symbolic relationships Type u Hierarchical l Parent / Child l Broader / Narrower than u Derived from hierarchies l Siblings (children of parents) u Associative l Other u Various flavors of near-synonymy l Similar l Source asserted synonymy l Possible synonymy PAR/CHD RB/RN SIB RO RL SY RQ

36. 36 Symbolic relationships Attribute u Hierarchical l isa (is-a-kind-of) l part-of u Associative l location-of l caused-by l treats l … u Cross-references (mapping)

37. Heart Concepts Metathesaurus 22 225 97 4 12 931 Esophagus Left Phrenic Nerve Heart Valves Fetal Heart Medias- tinum Saccular Viscus Angina Pectoris Cardiotonic Agents Tissue Donors Anatomical Structure Fully Formed Anatomical Structure Embryonic Structure Body Part, Organ or Organ Component Pharmacologic Substance Disease or Syndrome Population Group Semantic Types Semantic Network

38. UMLS Semantic Network

39. 39 Semantic Network u Semantic types (135) l tree structure l 2 major hierarchies n Entity –Physical Object –Conceptual Entity n Event –Activity –Phenomenon or Process

40. 40 Semantic Network u Semantic network relationships (54) l hierarchical (isa = is a kind of) n among types –Animal isa Organism –Enzyme isa Biologically Active Substance n among relations –treats isa affects l non-hierarchical n Sign or Symptom diagnoses Pathologic Function n Pharmacologic Substance treats Pathologic Function

41. 41 “Biologic Function” hierarchy (isa) Biologic Function Pathologic FunctionPhysiologic Function Disease or Syndrome Cell or Molecular Dysfunction Experimental Model of Disease Organism Function Organ or Tissue Function Cell Function Molecular Function Mental or Behavioral Dysfunction Neoplastic Process Mental Process Genetic Function

42. 42 Associative (non-isa) relationships Organism process of Embryonic Structure Anatomical Abnormality Congenital Abnormality Acquired Abnormality Fully Formed Anatomical Structure Anatomical Structure part of Organism Attribute property of Body Substance contains, produces conceptual part of evaluation of Body System conceptual part of Body Part, Organ or Organ Component part of Tissue part of Cell part of Cell Component Gene or Genome Body Space or Junction adjacent to location of evaluation of Finding Laboratory or Test Result Sign or Symptom Biologic Function Physiologic Function Pathologic Function Body Location or Region conceptual part of conceptual part of Injury or Poisoning disrupts co-occurs with

43. 43 Why a semantic network? u Semantic Types serve as high level categories assigned to Metathesaurus concepts, independently of their position in a hierarchy u A relationship between 2 Semantic Types (ST) is a possible link between 2 concepts that have been assigned to those STs l The relationship may or may not hold at the concept level l Other relationships may apply at the concept level

44. 44 Relationships can inherit semantics Semantic Network Metathesaurus Adrenal Cortex Adrenal Cortical hypofunction Disease or Syndrome Body Part, Organ, or Organ Component Pathologic Function isa Biologic Function isa Fully Formed Anatomical Structure isa location of

45. 45 UMLS Summary u Synonymous terms clustered into concepts u Unique identifier u Finer granularity u Broader scope u Additional hierarchical relationships u Semantic categorization

46. 46 Integrating subdomains Biomedical literature Biomedical literature MeSH Genome annotations Genome annotations GO Model organisms Model organisms NCBI Taxonomy Genetic knowledge bases OMIM Clinical repositories Clinical repositories SNOMED Other subdomains Other subdomains … Anatomy UWDA UMLS

47. 47 Integrating subdomains Biomedical literature Biomedical literature Genome annotations Genome annotations Model organisms Model organisms Genetic knowledge bases Clinical repositories Clinical repositories Other subdomains Other subdomains Anatomy

48. Information integration Genomics as an example

49. 49 NF2 Gene, protein, and disease Neurofibromatosis 2 is an autosomal dominant disease characterized by tumors called schwannomas involving the acoustic nerve, as well as other features. The disorder is caused by mutations of the NF2 gene resulting in absence or inactivation of the protein product. The protein product of NF2 is commonly called merlin (but also neurofibromin 2 and schwannomin) and functions as a tumor suppressor.

50. 50 Schwannoma (acoustic neuroma) http://www.mayoclinic.com

51. 51

52. 52 NF2 gene http://staff.washington.edu/timk/cyto/human/ http://www.ncbi.nlm.nih.gov/mapview/

53. 53

54. 54 Merlin u Synonyms l Neurofibromin 2 l Schwannomin l Schwannomerlin l Neurofibromatosis-2 u 10 isoforms u Annotations l Negative regulation of cell proliferation l Cytoskeleton l Plasma membrane

55. 55

56. 56 Neurofibromatosis 2 (Type II neurofibromatosis, Bilateral acoustic neurofibromatosis) C0027832 NF2 (Neurofibromin 2 gene) C0085114 Merlin (Schwannomin, Neurofibromin 2) C0254123 NEUROFIBROMATOSIS, TYPE II; NF2 #101000 Drosophila melanogaster merlin (Dmerlin) mRNA, complete cds. U49724 OMIMGenbank External resources UMLS Metathesaurus (Concepts and relations) Amino Acid, Peptide, or Protein Biologically Active Substance Neoplastic ProcessGene or Genome UMLS Semantic Network (Semantic Types) Merlin, Drosophila Tumor suppressor genes Benign neoplasms of cranial nerves Neuro- fibromatoses Tumor suppressor proteins

57. 57 Limitations u Genes not systematically represented l Most gene products and diseases are u Gene/Gene product-Disease relations l Not systematically represented l Not explicitly represented (e.g., co-occurrence) u Cross-references not systematically represented u Naming conventions (genes)

58. 58 References  UMLS umlsinfo.nlm.nih.gov u UMLS browsers (free, but UMLS license required) Knowledge Source Server: umlsks.nlm.nih.gov Knowledge Source Server: umlsks.nlm.nih.gov Semantic Navigator: http://mor.nlm.nih.gov/perl/semnav.pl Semantic Navigator: http://mor.nlm.nih.gov/perl/semnav.pl l RRF browser (standalone application distributed with the UMLS)

59. 59 References u Recent overviews l Bodenreider O. (2004). The Unified Medical Language System (UMLS): Integrating biomedical terminology. Nucleic Acids Research; D267-D270. l Nelson, S. J., Powell, T. & Humphreys, B. L. (2002 ). The Unified Medical Language System (UMLS) Project. In: Kent, Allen; Hall, Carolyn M., editors. Encyclopedia of Library and Information Science. New York: Marcel Dekker. p.369-378.

60. 60 References u UMLS as a research project l Lindberg, D. A., Humphreys, B. L., & McCray, A. T. (1993). The Unified Medical Language System. Methods Inf Med, 32(4), 281-91. l Humphreys, B. L., Lindberg, D. A., Schoolman, H. M., & Barnett, G. O. (1998). The Unified Medical Language System: an informatics research collaboration. J Am Med Inform Assoc, 5(1), 1-11.

61. 61 References u Technical papers l McCray, A. T., & Nelson, S. J. (1995). The representation of meaning in the UMLS. Methods Inf Med, 34(1-2), 193-201. l Bodenreider O. & McCray A. T. (2003). Exploring semantic groups through visual approaches. Journal of Biomedical Informatics, 36(6), 414-432.

62. UMLS in Use Mapping across Vocabularies

63. 63 The problem u For noun phrases extracted from medical texts, map to UMLS concepts u Then, select from the MeSH vocabulary the concepts that are the most closely related to the original concepts Medical text Noun phrase UMLS MeSH descriptor

64. 64 Map noun phrases to UMLS u Normalization l normalize noun phrases l use the normalized string index u MetaMap l approximate matching l more aggressive approach n use derivational variants n allow partial matches

65. 65 Restrict to MeSH u Based on the principle of semantic locality u Use different components of the UMLS u 4 techniques of increasing aggressiveness Use Synonymy MRCON + MRSO Use Synonymy MRCON + MRSO Use Associated expressions (ATXs) MRATX Use Associated expressions (ATXs) MRATX Explore the Ancestors MRREL + SN Explore the Ancestors MRREL + SN Explore the Other related concepts MRREL + SN Explore the Other related concepts MRREL + SN

66. 66 Restrict to MeSH: Synonymy u Term mapped to Source concept  For this concept, is there a synonym term that comes from MeSH? (MRSO)

67. 67 Restrict to MeSH: Assoc. expressions u If not,  Is there an associated expression (ATX) that describes this concept using a combination of MeSH descriptors? (MRATX) Endoscopic removal of intraluminal foreign body from oesophagus without incision AND Foreign Bodies MH/SH Esophagussurgery

68. 68 Restrict to MeSH: Ancestors u If not, let us build the graph of the ancestors of this concept using parents and broader concepts (MRREL) using parents and broader concepts (MRREL) l all the way to the top excluding ancestors whose semantic types are not compatible with those of the source concept (MRSTY) excluding ancestors whose semantic types are not compatible with those of the source concept (MRSTY)  From the graph, select the concepts that come from MeSH (MRCONSO) u Remove those that are ancestors of another concept coming from MeSH

69. 69 Restrict to MeSH: Other related concepts  If not, explore the other related concepts (MRREL) whose semantic types are compatible with those of the source concept (MRSTY)  From those, select the concepts that come from MeSH (MRCONSO)

70. 70 Restrict to MeSH: Example Vein of neck, NOS There is a MeSH term in the synonyms of SC SC is described by a combination of MeSH terms (ATX) The ancestors of SC contain MeSH terms MeSH terms from non-hierarchically related concepts Neck + Vein

71. 71 Restrict to MeSH: Example Vein of neck, NOS Vein of head and neck, NOS NeckBlood VesselsVascular structureVeins Systemic veins HeadHead and neck, NOSBody part, NOS

72. 72 Overall results u Synonymy:24% u Built-in mapping: 1% u Ancestors l From concept:49% l From children: 2% l From siblings: 1% u Other:11% u No mapping12%

73. 73 References u Bodenreider O, Nelson SJ, Hole WT, Chang HF. Beyond synonymy: exploiting the UMLS semantics in mapping vocabularies. Proceedings of AMIA Annual Symposium 1998:815-9. http://mor.nlm.nih.gov/pubs/pdf/1998-amia-ob.pdf http://mor.nlm.nih.gov/pubs/pdf/1998-amia-ob.pdf u Fung KW, Bodenreider O. Utilizing the UMLS for semantic mapping between terminologies. Proceedings of AMIA Annual Symposium 2005:266-270. http://mor.nlm.nih.gov/pubs/pdf/2005-amia-kwf.pdf http://mor.nlm.nih.gov/pubs/pdf/2005-amia-kwf.pdf

74. Advanced Library Services Towards a Biomedical Knowledge Repository

75. 75 Disclaimer u The project presented in this talk is being proposed as a new research initiative at the Lister Hill National Center for Biomedical Communications u It has not been approved or reviewed by NLM yet u The ideas presented here may not reflect NLM’s views u In collaboration with Tom Rindflesch, NLM

76. 76 Delivering Health Information u Provide biomedical text to health care professionals and consumers u Maintain NLM’s cutting edge l Support public health and healthy behavior l Assist clinical practice l Enable biomedical research and discovery u Exploit current Library resources and advanced technology

77. 77 Why additional services? u Biomedical literature is growing at an increasingly faster pace l High-throughput approach to literature processing u Integration between literature and other resources is insufficient l Adequate for navigating purposes l Insufficient for knowledge processing u Information retrieval is the starting point, not the end of the journey for the researcher

78. 78 Integration for navigation purposes http://www.ncbi.nlm.nih.gov/gquery/gquery.fcgi

79. 79 What additional services? u Multi-document summarization l Extract and visualize the facts extracted from 250 recent abstracts on the treatment of Parkinson’s disease u Question answering l Clinical and biological questions u Knowledge discovery l Connect facts from heterogeneous resources u Refined information retrieval l Indexing on relations in addition to concepts or association main heading/subheading

80. 80 Fact-based vs. concept-based u (concept, relationship, concept) triples are the common denominator to the various advanced services l Facts l Relations l Semantic predications l RDF triples

81. 81 Biomedical knowledge repository u Knowledge integration l Unique repository l Common format l Seamless environment l Phenotype and genotype information together u Enabling resource for the various services l Summarization l Question answering l Knowledge discovery l Refined information retrieval

82. 82 Sources of knowledge u Biomedical literature l Facts extracted from MEDLINE abstracts and full-text publicly available articles using text mining techniques l Other corpora u Structured databases / knowledge bases l NCBI resources l Model organism databases l Terminological knowledge l … u Contributed knowledge l The repository is open to collaborators outside NLM

83. 83 Annotated knowledge u Provenance information l Source (e.g., PMID) l Extraction mechanism l Timestamp u Frequency information l Redundancy u Collaborative annotation l “Was this information useful?” l Context of use/usefulness

84. 84 Semantic Web perspective u Common format for knowledge l Resource Description Format (RDF) u Common identification scheme l Unified Resource Identifier (URI) u Standard tools l RDF browsers l RDF “reasoners” u High level of interest for biomedicine in the SW community l Health Care and Life Sciences Interest Group

85. Biomedical Knowledge Repository Information Retrieval Question Answering Knowledge Discovery Document Summarization Source selection Text Mining Terminological Knowledge Other K. Sources Contributed Knowledge UMLS MetaMap MEDLINECT.gov Model organism annotation databases OMIM SemRep 01DEC05

86. Towards a Biomedical Knowledge Repository

87. 87 Creating the repository Biomedical Knowledge Repository Medline Structured Biomedical Data Enhanced Information Management for Medicine Text Mining PubMed Clinical Trials.gov UMLS Contributed Knowledge

88. 88 Creating the repository Structured Biomedical Data PubMed Enhanced Information Management for Medicine UMLS Contributed Knowledge Biomedical Knowledge Repository Medline Text Mining Semantic relations e.g., Rasagiline TREATS Parkinson Disease Clinical Trials.gov

89. 89 Creating the repository Medline PubMed Enhanced Information Management for Medicine Text Mining Clinical Trials.gov Biomedical Knowledge Repository Structured Biomedical Data UMLS Contributed Knowledge

90. 90 Contributed Knowledge UMLS Creating the repository Medline PubMed Enhanced Information Management for Medicine Text Mining Clinical Trials.gov Biomedical Knowledge Repository Structured Biomedical Data

91. 91 Contributed Knowledge UMLS Creating the repository Medline PubMed Enhanced Information Management for Medicine Text Mining Clinical Trials.gov Biomedical Knowledge Repository Entrez Gene

92. 92 Contributed Knowledge UMLS Creating the repository Medline PubMed Enhanced Information Management for Medicine Text Mining Clinical Trials.gov Biomedical Knowledge Repository Genetics Home Reference

93. 93 Contributed Knowledge UMLS Creating the repository Medline PubMed Enhanced Information Management for Medicine Text Mining Clinical Trials.gov Biomedical Knowledge Repository OMIM

94. 94 Contributed Knowledge UMLS Creating the repository Medline PubMed Enhanced Information Management for Medicine Text Mining Clinical Trials.gov Biomedical Knowledge Repository Gene Ontology

95. 95 Structured Biomedical Data UMLS Creating the repository Medline PubMed Enhanced Information Management for Medicine Text Mining Clinical Trials.gov Biomedical Knowledge Repository Contributed Knowledge

96. 96 Contributed Knowledge Structured Biomedical Data UMLS Advanced library services Medline PubMed Text Mining Clinical Trials.gov Biomedical Knowledge Repository Enhanced Information Management for Medicine

97. 97 Contributed Knowledge Structured Biomedical Data UMLS Advanced library services Medline PubMed Text Mining Clinical Trials.gov Biomedical Knowledge Repository Knowledge Discovery

98. 98 Contributed Knowledge Structured Biomedical Data UMLS Advanced library services Medline PubMed Text Mining Clinical Trials.gov Biomedical Knowledge Repository Question Answering

99. 99 Contributed Knowledge Structured Biomedical Data UMLS Advanced library services Medline PubMed Text Mining Clinical Trials.gov Biomedical Knowledge Repository Summarization and Focused Retrieval

100. Summarizing Biomedical Text

101. 101 Summarizing Biomedical Text u Search l Medline l ClinicalTrials.gov u Summarize documents l Most salient semantic relations u Visualize the summary u Link the semantic relations to l Original text l Related structured knowledge

102. 102 Text Mining Workflow Information retrieval summarization Text mining Parkinson disease/therapy Network of relations retrieval 294 articles

103. 103 Text Mining Workflow Parkinson disease/therapy

104. 104 Movement Disorders Parkinson Disease pramipexol Dopamin Agonists Dopamine Brain rasagilineLevodopa Entire subthalamic nucleus Neuro- degenerative Diseases entacapone Anhedonia treats location of Gene Therapy Deep brain Stimulation Procedure Depressive disorder Bilateral breast cancer Dementia occurs in Dyskinetic syndrome isa treats Treatment of Parkinson’s disease SemGen output

105. 105 Movement Disorders Parkinson Disease pramipexol Dopamin Agonists Dopamine Brain rasagilineLevodopa Entire subthalamic nucleus Neuro- degenerative Diseases entacapone Catechol-O-methyl- transferase inhibitor Anhedonia Monoamine Oxidase Inhibitors Antiparkinson Agents Antidepressive Agents treats isa location of part of Gene Therapy Deep brain Stimulation Procedure Depressive disorder Bilateral breast cancer Dementia occurs in Dyskinetic syndrome isa treats associated with SemGen output + UMLS relations + additional UMLS concepts SemGen output + UMLS relations + additional UMLS concepts Treatment of Parkinson’s disease

106. 106 Conclusions u Need to go beyond information retrieval u Need to integrate multiple, heterogeneous knowledge sources to support knowledge processing, not only navigation u Synergistic with the Semantic Web l Emerging standard framework l W3C Health Care and Life Sciences Interest Group http://www.w3.org/2001/sw/hcls/ http://www.w3.org/2001/sw/hcls/

107. Medical Ontology Research Olivier Bodenreider Lister Hill National Center for Biomedical Communications Bethesda, Maryland - USA Contact:Web:olivier@nlm.nih.govmor.nlm.nih.gov