louvainCluster: Louvain algorithm for community detection

In rliger: Linked Inference of Genomic Experimental Relationships

Louvain algorithm for community detection

Description

After quantile normalization, users can additionally run the Louvain algorithm for community detection, which is widely used in single-cell analysis and excels at merging small clusters into broad cell classes.

Usage

louvainCluster(
  object,
  resolution = 1,
  k = 20,
  prune = 1/15,
  eps = 0.1,
  nRandomStarts = 10,
  nIterations = 100,
  random.seed = 1,
  verbose = TRUE,
  dims.use = NULL
)

Arguments

object	liger object. Should run quantile_norm before calling.
resolution	Value of the resolution parameter, use a value above (below) 1.0 if you want to obtain a larger (smaller) number of communities. (default 1.0)
k	The maximum number of nearest neighbours to compute. (default 20)
prune	Sets the cutoff for acceptable Jaccard index when computing the neighborhood overlap for the SNN construction. Any edges with values less than or equal to this will be set to 0 and removed from the SNN graph. Essentially sets the strigency of pruning (0 — no pruning, 1 — prune everything). (default 1/15)
eps	The error bound of the nearest neighbor search. (default 0.1)
nRandomStarts	Number of random starts. (default 10)
nIterations	Maximal number of iterations per random start. (default 100)
random.seed	Seed of the random number generator. (default 1)
verbose	Print messages (TRUE by default)
dims.use	Indices of factors to use for Louvain clustering (default 1:ncol(H[[1]])).

Value

liger object with refined 'clusters' slot set.

Examples

ligerex <- createLiger(list(ctrl = ctrl, stim = stim))
ligerex <- normalize(ligerex)
ligerex <- selectGenes(ligerex)
ligerex <- scaleNotCenter(ligerex)
ligerex <- optimizeALS(ligerex, k = 5, max.iters = 1)
ligerex <- quantile_norm(ligerex)
ligerex <- louvainCluster(ligerex, resolution = 0.3)

rliger documentation built on Nov. 9, 2023, 1:07 a.m.