NEMO: NEMO: neighborhood-based multi-omics clustering

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

NEMO: neighborhood-based multi-omics clustering

Description

NEMO \insertCiteWang2014aMosaiClusteR integrates several omics by building a locally-scaled, k-nearest-neighbour affinity network for each one and averaging them, for every pair of objects, over only the omics in which both objects are measured. Partial / missing modalities are therefore handled without imputation. The integrated affinity is clustered spectrally.

Usage

NEMO(List, k = NULL, NN = 20, normalize = TRUE)

Arguments

List	A list of data matrices over the same objects, objects in rows. Sources may cover different subsets of objects (matched by row name) for the partial-data setting.
k	Number of clusters (NULL = eigengap estimate).
NN	Number of neighbours for the affinity graphs and local scaling.
normalize	Logical; row-normalise each affinity to a transition matrix before integration (NEMO default TRUE).

Value

A list of class "NEMO" with FusedM (integrated affinity), DistM (1 - FusedM), cluster (labels) and Clust (an agnes object on DistM for interoperability).

References

\insertCite

Wang2014aMosaiClusteR; Rappoport, N. & Shamir, R. (2019). NEMO: cancer subtyping by integration of partial multi-omic data. Bioinformatics 35:3348-3356.

See Also

SNF, spectral_clustering

Examples

data(mosaic_toy)
fit <- NEMO(mosaic_toy$List, k = 3, NN = 15)
table(fit$cluster)

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