Date of Award

2018

Publisher

North Dakota State University

Document Type

Thesis

Degree Awarded

Master of Science (MS)

Program

Computer Science

Department

Computer Science

College

Engineering

Faculty Advisor

Salem, Saeed

Description

Frequent graph mining has received considerable attention from researchers. Existing algorithms for frequent subgraph mining do not scale for large networks, and take hours to finish. Mining multiple gene coexpressions networks allows for identifying context-specific modules. Frequent subnetworks represent essential biological modules. In this thesis, we propose two algorithms for mining frequent subgraphs. In the first algorithm, we propose a parallel algorithm for mining maximal frequent subgraphs from gene coexpression networks. Despite the algorithm’s parallelization, it takes much time and it does not allow relaxation. This inspired us to develop a second algorithm that solves those problems. In the second algorithm, we propose a greedy approach for mining approximate frequent subgraphs. Experiments on real tissue-specific RNA-seq expression networks and synthetic data demonstrate the effectiveness of the proposed algorithms. Moreover, biological enrichment analysis shows that the reported patterns are biologically relevant and enriched with known biological processes and KEGG pathways.

Rights

NDSU Policy 190.6.2

Rights Link

https://www.ndsu.edu/fileadmin/policy/190.pdf

Author ORCID Identifier

0000-0002-1473-5070

Handle Identifier

https://hdl.handle.net/10365/28735

Download

Share

COinS