What multi-omics integration methods are available?

OmicsAnalyst
1

OmicsAnalyst 2.0 offers multiple integration strategies:

Unsupervised Methods:

  • Consensus PCA - Identify common patterns across omics layers

  • MCIA - Multiple co-inertia analysis for global exploration

  • MOFA - Multi-Omics Factor Analysis for latent factor discovery

  • Fast NMF - Non-negative matrix factorization for pattern discovery

Supervised Methods:

  • DIABLO - Multi-block discriminant analysis for biomarker discovery and feature selection

Pairwise Methods:

  • Sparse CCA - Canonical correlation analysis for two-layer integration

All methods can export top-loading features to OmicsNet for knowledge-based regulatory network integration and functional interpretation.