Model Formalization for Genomes Comparative Analysis Using a Graph Database

The paper aims to discuss the problem of graph data model formalization for genomic information representation. The model being discussed is designated for the implementation into the graph database for comparative genomics tasks. The paper focuses on the task of homology search, which may be defined as discovering DNA regions reflecting shared ancestry between host organisms usually inferred from the excess sequences similarity. Thus, homology search implies taking stochastic DNA mutations into account while evaluating sequences similarity. The proposed graph theory-based model aims to represent both probabilistic and deterministic aspects of the genomes comparison task by implementing a de Bruijn Graph structure along with an underlying Markov chain.