Identification of fever and vaccine- associated gene interaction - - PowerPoint PPT Presentation
Identification of fever and vaccine- associated gene interaction networks using ontology-based literature mining Arzucan zgr Bogazici University Junguk Hur, Zuoshuang Xiang, and Yongqun Oliver He University of Michigan The VDOSME
Bogazici University
University of Michigan The VDOSME workshop, ICBO 2012 July 21, 2012
Fever is a symptom of abnormal elevation of body
Fever-associated genes include PGE2, PLA2, COX-2,
Many vaccines cause fever, but how vaccination
Goal: Identify gene-gene and gene-vaccine
Fever-related articles obtained from PubMed:
“Fever OR Hyperthermia OR Pyrexia OR Febrile OR Pyrexial”
Vaccine and fever-related articles:
including the terms “vaccine”, “vaccination”, and their variants
including 186 specific vaccine names from VO
Sentences of titles and abstracts obtained from:
BioNLP database in the National Center for Integrative
Ontology of the vaccine domain for vaccine data
http://www.violinet.org/vaccineontology/
Classifies a large number of existing vaccines (>
Follows the OBO Foundry principles. Led by Yongqun “Oliver” He (co-author of this
Gene names tagged using SciMiner (Hur et al.,
Dictionary and rule-based system F-score: 76%
Genes reported in terms of the official human genes
VO-SciMiner used to identify vaccine names based on a
F-score: 95%
IL-2 and IL-15 induced the production of IL-17 and IFN-gamma in a dose dependent manner by PBMCs. No interaction. No interaction. interaction Path between proteins: good description of semantic relation between them.
Stanford Parser is used to generate the dependency parse trees (de Marneffe et al., 2006).
Minimum number of operations (insertion, deletion, or substitution of a single
IL2 – nsubj – induced – dobj – production – prep_of – IL-17 IL2 – nsubj – induced – dobj – production – prep_of – IL-17 – conj_and – IFN-gamma IL-17 – conj_and – IFN-gamma
Edit distance (Path1 -> Path2) = 2 (2 insertions) Edit distance (Path1 -> Path3) = 8 (6 deletions + 2 insertions)
Convert to Similarity Function:
2007; Ozgur et al., Journal of Biomedical Semantics, 2011).
−γEditDistpi,p j]
The number of nodes a given node is connected to Measures the extent of inluence a node has on the network The more neighbors a node has, the more important it is Degree centrality of x = 5; of y = 2
x z
j=1 n
Proportional to the sum of the centralities of the neighbors of
In matrix representation: λx = Ax For non-negative centrality vector: λ is largest eigenvalue of A and x is the corresponding eigenvector Not all neighbors contribute equally to the centrality of node Defined as “prestige” in social networks The prestige of a person depends not only on how many
j= 1 n
Inverse sum of the geodesic distances from a given node to
The closer a node to the other nodes, the more important it is.
x Geodesic distance: length of shortest path between node i and node j (dij)
j= 1 n
−1
For a node i: sum over all pairs of nodes of proportion of the
Control of a node over the information flow of the network A node is important if it is on many geodesics
x gjk(i): # of geodesics passing over i gjk: total # of geodesics
j<k
The Database for Annotation, Visualization and
997 significantly over-represented functional terms (GO
239 significantly over-represented functional terms (GO
New scientific hypothesis can be generated (e.g. Role of
SVM pipeline applied to extract gene-vaccine
1,716 articles containing 2,835 interactions identified.
New scientific hypothesis can be generated.
Identification of fever and fever–vaccine associated
Improved mining performance by VO Phosphorylation-focused regulation enriched in the
Identification of TLRs as potential key factors in
Expansion of the networks by
including more specific vaccines (via improved VO) using a sentence-level co-citation rather than SVM-based
integrating Ontology of Adverse Events (OAE;
Apply the ontology-based literature mining approach
Junguk Hur Yongqun He Zoushuang Xiang Rebecca Racz Dragomir Radev Eva Feldman Alex Ade Brian Athey