cameron.bibm2011

Semantic Predications for Complex Information Needs in Biomedical Literature

Delroy Cameron, Ramakanth Kavuluru, Pablo N. MendesAmit P. Sheth, Krishnaprasad Thirunarayan

Ohio Center for Excellence in Knowledge-enabled Computing kno.e.sis Center, Wright State University

Dayton, OH 45435, USA

Olivier BodenreiderNational Library of Medicine Bethesda, MD 20894, USA

48th ACM Southeast Conference. ACMSE 2010. Oxford, Mississippi. April 15-17, 2010.

2011 International Conference on Bioinformatics and Biomedicine (BIBM)12-15 November, 2011 Atlanta, Georgia

MOTIVATION

2

• Information Retrieval Interaction Sequence

– Keyword Search– Document Selection– Document Inspection– Query Reformulation

Document-Centric Model– Hyperlink-driven Browsing– Information is within Documents

Limitations– Query Reformulation– Constrained Navigation

Exploratory Search

“How do mutations in the Presenilin-1 (PS1) gene affect Alzheimer’s disease (AD)?”

. . . mutations in PS1 lead to Alzheimer’s disease by increasing the extracellular levels of [amyloid peptide 42] A42. (Source: PMID10652366)

. . . familial early onset Alzheimer’s disease is caused by point mutations in the amyloid precursor protein gene on chromosome 21, in the presenilin 2(PS2)1 gene on chromosome 1, or, most frequently, in the presenilin 1(PS1) gene on chromosome14. . . (Source:PMID9013610)chromosome14 finding_site_of presenilin1

PS1 associated_with Alzheimer’s Disease

Semantic Predications

COMPLEX INFO NEEDS

Literature-Based Discovery (LBD)Don R. Swanson’s Hypotheses

o Raynaud’s Syndrome-Dietary Fish Oilo Magnesium-Migraine

Question AnsweringText REtrieval Conference (TREC)2006, 28 questions

6

TREC Genomics Track - http://ir.ohsu.edu/genomics/

PROBLEM

COMPLEX BIOMEDICAL QUESTION

7

ANSWER DOCUMENTS o LEGAL SPANS

SEMANTIC PREDICATIONS-BASED RETRIEVAL

REACHABILITY

8

“the notion of being able to get from one vertex in a directed graph to some other vertex”

a

c

bMagnesium

Calcium Channel Blocker

ISA

INVERSE_ISA

Labeled Graph

REACHABILITY-DOCS

9

“is the notion of being able to cover the documents in a document set, using the vertices in a directed graph from one vertex to some other vertex”

Predications Graph Plane

Document Plane

11

• Stopping Conditions– No Reachable Docs in PG– No Successors in PG +Reachable Docs in DP– No Successors in PG + No Reachable Docs

in DP

Knowledge Abstraction

Predications Graph Plane

Document Plane

C0040682 - cell transformation

C1261468 - Cell fusion

C0007613 – Cell physiology

coexists_with

coexists_with

12

Reachability Framework

Predications Graph (PG) Plane

Document Plane

External Knowledge Base Graph Plane

Provenance

Knowledge Abstraction

DATASET

13

• TREC 2006 Corpus 26 Questions 162,259 full text documents 12,641,116 text items

Gold Standard 1381 Gold Standard Documents 3461 Text items

• Biomedical Knowledge Repository (BKR) 13 million from UMLS Metathesaurus 8 million from Literature using SemRep

TREC Genomics Track - http://ir.ohsu.edu/genomics/

EXPERIMENTS

14

• Experiment I: Single-Graph 2105 Vertices, 16942 Edges > 13,000 unique predications 240 no predications, 3 not processed 121,162 text items

Experiment II: Multiple-Graph 26 predications graph

OBSERVATIONS

17

Absence of predications in text Predication extraction methods (SemRep) Absence of direct connections among text item Ambiguity in written language Abstractions may lead to information overflow Quality of background knowledge

“How do mutations in the Pes gene affect cell growth?”

G1_phase G2_phase

DNA Replication

• Novel Knowledge-driven framework • Semantic Predications to link scientific content• Alternative to Query reformulation• Background knowledge boost recall

• Effective at Coarse granularity• Poor at fine granularity

CONCLUSION

19

• Path/Predication Ranking• Path/Predication Collapsing• Knowledge Abstraction tuning• Scalability • Computational Complexity

• Replicate Swanson’s Hypotheses

FUTURE WORK

20

ACKNOWLEDGEMENT

National Library of Medicine (NIH/NLM)

Human Performance & Cognition Ontology Project @knoesis

Cartic Ramakrishnan Michael Cooney Gary Alan Smith Paul Fultz II Jeffrey Ali Hyacinthe Thomas C. Rindflesch Mohamed Cyclegar Dongwook Shin John Nguyen May Cheh

21

QUESTIONS

22

Topic ID Question

160 What is the role of PrnP in mad cow disease?

161 What is the role of IDE in Alzheimer's disease

163 What is the role of APC (adenomatous polyposis coli) in colon cancer?

165 How do Cathepsin D (CTSD) and apolipoprotein E (ApoE) interactions contribute to Alzheimer's disease?

167 How does nucleoside diphosphate kinase (NM23) contribute to tumor progression?

168 How does BARD1 regulate BRCA1 activity?

169 How does APC (adenomatous polyposis coli) protein affect actin assembly

170 How does COP2 contribute to CFTR export from the endoplasmic reticulum?

172 How does p53 affect apoptosis?

173 How do alpha7 nicotinic receptor subunits affect ethanol metabolism?

174 How does BRCA1 ubiquitinating activity contribute to cancer?

176 How does Sec61-mediated CFTR degradation contribute to cystic fibrosis?

177 How do Bop-Pes interactions affect cell growth?

178 How do interactions between insulin-like GFs and the insulin receptor affect skin biology?

179 How do interactions between HNF4 and COUP-TF1 suppress liver function?

180 How do Ret-GDNF interactions affect liver development?

184 How do mutations in the Pes gene affect cell growth?

185 How do mutations in the hypocretin receptor 2 gene affect narcolepsy?

186 How do mutations in the Presenilin-1 gene affect Alzheimer's disease?

TREC Genomics Track 2006 (Questions)

Most Relevant Subtreea

Level-1

Level-2

Level-3

Level-4

Level-0