Table 1 Broad categorization of graph machine learning techniques for multi-omics data (adapted from Xia et al. [19]) with literature examples.
From: Graph machine learning for integrated multi-omics analysis
References | |
|---|---|
Methods based on random walks | |
They are useful for node classification and graph clustering. They simulate a process where a walker moves from one node to another in the graph by following edges randomly. | [43] |
Methods based on matrix factorization | |
They involve decomposing matrices associated with graphs into the product of two or more matrices, and are employed in miscellaneous graph-based learning tasks. | [44] |
Methods based on deep learning | |
They can learn representations and features from graph-structured data, e.g., graph autoencoders, graph convolutional networks, graph attention networks, and temporal graph networks. | |
