Multimodal Methods for Knowledge Discovery from Bulk and Single-Cell Multi-Omics Data

Multi-omics measurements (bulk and single-cell) are essential to depict the cellular states comprehensively and thus could derive a deep understanding of underlying mechanisms for cellular state changes in many biological processes. Therefore, computational models that integrate omics data are often indispensable for discovering novel effective diagnostics and therapeutics. In this chapter, we will give an overview of the multimodal methods for a variety of data analysis tasks (dimensionality reduction, clustering, gene regulatory network inference, and biomarker discovery), leading to the discovery of cell populations, gene regulatory networks, and biomarkers. We will also cover the characteristics associated with each method to provide users with practical guidance on how to choose appropriate methods for their specific application scenarios.