mofaplus-multi-omics

You integrate scRNA-seq and ATAC-seq data using Multi-Omics Factor Analysis v2 (MOFA+) to perform joint decomposition, identifying shared and view-specific latent factors that represent coordinated biological variation across modalities.

Multi-omics factor analysis is a joint decomposition technique that reduces complex multi-omics single-cell data into interpretable latent factors. You need it when discovering hidden biological drivers spanning across RNA, ATAC, protein, or methylation modalities.

Yes, MOFA+ supports multi-group experimental designs for comparative analysis. You can integrate multiple omics layers across different sample groups to identify both shared and view-specific latent factors representing biological variation.

You need AnnData objects for your omics layers, such as scRNA-seq and ATAC-seq data. The trained MOFA+ model output is saved as an HDF5 file, requiring h5py alongside mofapy2 and muon dependencies.

Yes, mofapy2 works with muon to perform multi-omics dimensionality reduction. This combination integrates multiple biological measurements into a reduced set of interpretable latent factors representing underlying biological variation.