Using Hypergraphs to Model Complex Biological Networks, a Review on Challenges and Opportunities.

Abstract

Traditionally, Relational DBMSs were the method of choice for modelling any kind of data, however due to their inefficiencies in modelling big interconnected data, in the last two decades, Graph DBMSs attracted the interest of many researchers and scientists. In graph data modelling, nodes represent entities of the system and binary edges show the relationship between these entities. Although graph methods are powerful tools in theory and practice to model big data, they suffer from the limits in their relationship modelling due to the binary edges restriction. In biology systems and other complex systems, higher order relationships abound. Hypergraphs are the mathematical objects that are able to model higher order relationships, using multiway edges known as hyperedges. Hyperedges encode the relationship between 2 or more than 2 nodes in a hypergraph. In this work, we survey the usage of Hypergraphs as a novel method to model biological networks. We first explain complex systems theory, and illustrate why biologic systems are complex. Afterward, we highlight the new discoveries in the literature on how hypergraphs are of paramount importance to tackle this complexity. We review the challenges and opportunities that hypergraph modelling introduced to computational biology and systems biology and discuss how different innovative strategies are used by different researchers in order to increase the complexity level of their method to cover the natural complexity of the problems under investigation. Finally, we discuss the possible future lines and current restrictions. Our main aim is to increase the awareness of hypergraph modelling. Most of the studies show hypergraphs are powerful mathematical objects to model complex biological systems.