Projects per year

## Abstract

Graph networks can model the data observed across different levels of biological systems that span from the population graph (with patients as network nodes) to the molecular graphs that involve omics data. Graph-based approaches have shed light on decoding biological processes modulated by complex interactions. This paper systematically reviews the graph-based analysis methods, including Graph Signal Processing (GSP), Graph Neural Network (GNN), and graph topology inference methods, and their applications to biological data. This work focuses on the algorithms of the graph-based approaches and the constructions of the graph-based frameworks that are adapted to the broad range of biological data. We cover the Graph Fourier Transform and the graph filter developed in GSP, which provides tools to investigate biological networks in the graph domain that can potentially benefit from the underlying graph structure. We also review the node, graph, and interaction oriented GNN architecture with inductive and transductive learning manners for various biological objectives. As the key component of graph analysis, we provide a review of the graph topology inference methods that incorporate assumptions for specific biological objectives. Finally, we discuss the biological application of graph analysis methods within the exhaustive literature collection, potentially providing insights for future research in the biological sciences.

Original language | English |
---|---|

Pages (from-to) | 109-135 |

Number of pages | 27 |

Journal | IEEE Reviews in Biomedical Engineering |

Volume | 16 |

DOIs | |

Publication status | Published - 2023 |

## Keywords

- biological data
- Biology
- Fourier transforms
- graph convolutional network
- graph learning
- graph neural network
- Graph signal processing
- Laplace equations
- Matrix decomposition
- Network topology
- Signal processing
- Topology

## Projects

- 1 Finished