High-throughput molecular profiling technologies have produced high-dimensional multi-omics data, enabling systematic understanding of living systems at the genome scale. Studying molecular interactions across different data types helps reveal signal transduction mechanisms across different classes of molecules. In this paper, we develop a novel Bayesian representation learning method that infers the relational interactions across multi-omics data types. Our method, Bayesian Relational Learning (BayReL) for multi-omics data integration, takes advantage of a priori known relationships among the same class of molecules, modeled as a graph at each corresponding view, to learn view-specific latent variables as well as a multi-partite graph that encodes the interactions across views. Our experiments on several real-world datasets demonstrate enhanced performance of BayReL in inferring meaningful interactions compared to existing baselines.

To further clarify the model and workflow of our proposed BayReL, we provide a schematic illustration of BayReL in Figure S1, where we only include two views for clarity. Figure S2 shows the inferred bipartite network with the top 200 interactions by BayReL. Schematic illustration of BayReL. 2 Figure S2: The bipartite sub-network with the top 200 interactions inferred by BayReL in AML data, Genes and drugs are shown as blue and red nodes, respectively. D. Details on the experimental setups, hyper-parameter selection, and run time We learn the model for 1000 training epochs and use the validation set for early stopping. Each training epoch for CF, BRCA, and AML took 0.01, 0.42, In all experiments, we used CCAGFA R package as the official implementation of BCCA.

Summary and Contributions: In this paper, the authors propose a Bayesian representation learning framework that can infer links between heterogeneous graphs generated from multi-omics datasets. The main idea is to use the underlying relationship information within each dataset (or view) by modeling it as a graph. The method has 4 steps - (1) to embed the nodes of each view-specific graph into in the same latent space (2) generate a multi-view adjacency tensor using the similarity scores for node embeddings across views (3) Infer prior latent variables from the node embeddings and multi-view graphs and posterior from the view-specific data (4) Finally, perform variational inference to optimize model parameters and variational parameters. The paper attempts to solve an important problem of multi-omics data integration by learning relationships that can exist between different modalities by modeling them as multi-view link prediction. This work could be useful to the broader ML community.

The paper proposes a Bayesian formulation for the integration of multi omics datasets by combining within-view and between-view interactions. Although the paper is conceptually related to prior work, the reviewers appreciate the contributions made, which are both timely and relevant to the neurips community. Overall, this is a solid submission and the authors defend the concerns raised convincingly in their rebuttal.

High-throughput molecular profiling technologies have produced high-dimensional multi-omics data, enabling systematic understanding of living systems at the genome scale. Studying molecular interactions across different data types helps reveal signal transduction mechanisms across different classes of molecules. In this paper, we develop a novel Bayesian representation learning method that infers the relational interactions across multi-omics data types. Our method, Bayesian Relational Learning (BayReL) for multi-omics data integration, takes advantage of a priori known relationships among the same class of molecules, modeled as a graph at each corresponding view, to learn view-specific latent variables as well as a multi-partite graph that encodes the interactions across views. Our experiments on several real-world datasets demonstrate enhanced performance of BayReL in inferring meaningful interactions compared to existing baselines.