Analysis of Biologic Relationships with Graph Databases

Advisor Information

Dhundy Bastola

Location

UNO Criss Library, Room 249

Presentation Type

Oral Presentation

Start Date

6-3-2015 10:15 AM

End Date

6-3-2015 10:30 AM

Abstract

Biologic data and medical data are replete with examples of inter-related pieces of information. Proteins interact with other proteins, proteins are involved in metabolic relationships, genes are involved in disease relationships; genes have relationships with function and structure; genes have expression level relationships with tissues which helps to explain tissue level differences. Medical terminology is encoded in ontologies where the relationship of a clinical diagnosis is made to all of the factors which define the diagnosis. The common thread is that relationships are established and documented. One method of analysis of relationships is the creation of networks which represent these relationships. A computational representation of relationships can be created in a graph databases which store nodes and their relationships. The rich library of existing Graph Theory algorithms can be applied to networks to extract information from them. Several projects were undertaken in the previous year in the UNO Bioinformatics lab relating to gene expression analysis, metabolic network analysis and representing a clinical ontology of 300,000 terms in a network. The analysis of the network of metabolic reactions used by organisms found to reside in the microbiome of the mouth was undertaken. A possible connection between the disease gingivitis and the metabolic networks of organisms in the mouth was found and published. Ongoing research continues examining the microbiome of the GI tract which will be compared to that of the mouth.

This document is currently not available here.

COinS
 
Mar 6th, 10:15 AM Mar 6th, 10:30 AM

Analysis of Biologic Relationships with Graph Databases

UNO Criss Library, Room 249

Biologic data and medical data are replete with examples of inter-related pieces of information. Proteins interact with other proteins, proteins are involved in metabolic relationships, genes are involved in disease relationships; genes have relationships with function and structure; genes have expression level relationships with tissues which helps to explain tissue level differences. Medical terminology is encoded in ontologies where the relationship of a clinical diagnosis is made to all of the factors which define the diagnosis. The common thread is that relationships are established and documented. One method of analysis of relationships is the creation of networks which represent these relationships. A computational representation of relationships can be created in a graph databases which store nodes and their relationships. The rich library of existing Graph Theory algorithms can be applied to networks to extract information from them. Several projects were undertaken in the previous year in the UNO Bioinformatics lab relating to gene expression analysis, metabolic network analysis and representing a clinical ontology of 300,000 terms in a network. The analysis of the network of metabolic reactions used by organisms found to reside in the microbiome of the mouth was undertaken. A possible connection between the disease gingivitis and the metabolic networks of organisms in the mouth was found and published. Ongoing research continues examining the microbiome of the GI tract which will be compared to that of the mouth.