MosaiClusteR: MosaiClusteR: An Umbrella Framework for Multi-Source and...

In MosaiClusteR: An Umbrella Framework for Multi-Source and Multi-Omics Clustering

MosaiClusteR: An Umbrella Framework for Multi-Source and Multi-Omics Clustering

Description

MosaiClusteR ("MoSaIC" = Multi-Omics Source-Agnostic Integration Clustering in R) unifies a large family of multi-source clustering methodologies behind a single, consistent list-of-matrices interface, and wraps them in a complete analysis framework: preprocessing, single-source baselines, integrative clustering across five paradigms, and downstream comparison/evaluation.

The MoSaIC framework (how to use the package)

1. Preprocess each source with Normalization and Distance.

2. Baseline each source on its own (any hierarchical clusterer) to motivate integration.

3. Integrate with one or more methods, e.g. M_ABC, M_ABCdist, WeightedClust, SNF, CEC, HierarchicalEnsembleClustering.

4. Evaluate and compare solutions with compare_clusterings and cluster_agreement.

Every clustering method returns a list containing at least DistM (a dissimilarity matrix) and Clust (a hierarchical-clustering object), so results are directly comparable and composable.

Feature weighting with data nuggets

The M-ABC family weights features for subsampling. Beyond the classic variance ("var") and coefficient-of-variation ("cv") schemes, MosaiClusteR adds weighting = "nugget", a robust, Big-Data-friendly alternative based on create_data_nuggets. See nugget_feature_weights and ABCpp.SingleInMultiple.

Author(s)

Maintainer: Bernard Isekah Osang'ir [e-mail] (ORCID)

Authors:

• Bernard Isekah Osang'ir [e-mail] (ORCID)

• Marijke Van Moerbeke

Other contributors:

• Ziv Shkedy [contributor]

• Surya Gupta [contributor]

• Jürgen Claesen [contributor]

See Also

Useful links:

• https://github.com/lucp12891/MosaiClusteR

• Report bugs at https://github.com/lucp12891/MosaiClusteR/issues

MosaiClusteR documentation built on Oct. 25, 2026, 1:08 a.m.