bio-multi-omics-mixomics-analysis

You can integrate multi-omics datasets for classification by building DIABLO models with block.splsda to find discriminant multi-omics signatures and biomarker panels that separate experimental groups.

DIABLO is a supervised multi-block integration method used to identify cross-omics biomarker panels that discriminate between conditions, eliminating ad-hoc per-omics analyses and manual cross-omics comparison.

Yes, you can use sPLS-DA for single-omics classification, sPLS for pairwise correlated feature discovery, and spca for sparse dimensionality reduction on single-omics datasets.

Tune keepX via cross-validation, assess model performance using the perf function and auroc, and export the selected variables to validate discriminant features for downstream analysis.

You can generate consensus and per-block sample plots, circos plots, correlation networks, and heatmaps to visualize cross-omics feature correlations and sample separations.

Yes, MINT supports multi-study integration for transcriptomics, proteomics, and metabolomics datasets, allowing you to classify samples and identify discriminant features across different studies.