assess.cluster.similarity: Assess Cluster Similarity

In robustSingleCell: Robust Clustering of Single Cell RNA-Seq Data

Description

Assess similarity between pairs of clusters.

Usage

assess.cluster.similarity(environment,
  diff.exp.file = "main.datasets.diff.exp.rds",
  cluster.similarity.function = pearson.correlation, label = "pearson",
  rerun = F)

Arguments

environment	environment object
diff.exp.file	name of differential expression results file
cluster.similarity.function	which similarity function to use (either 'pearson.correlation' or '?') Mamie - there was another similarity function using euclidean distance. Do you know where did it go to? Can you replace the '?' with the name of this other function?
label	name of the similarity measure to use for the results folder
rerun	whether to rerun the analysis or load from cache

Value

pairwise cluster similarity measures

Examples

LCMV1 <- setup_LCMV_example()
LCMV1 <- get.variable.genes(LCMV1, min.mean = 0.1, min.frac.cells = 0,
min.dispersion.scaled = 0.1)
LCMV1 <- PCA(LCMV1)
LCMV1 <- cluster.analysis(LCMV1)
types = rbind(
data.frame(type='Tfh',gene=c('Tcf7','Cxcr5','Bcl6')),
data.frame(type='Th1',gene=c('Cxcr6','Ifng','Tbx21')),
data.frame(type='Tcmp',gene=c('Ccr7','Bcl2','Tcf7')),
data.frame(type='Treg',gene=c('Foxp3','Il2ra')),
data.frame(type='Tmem',gene=c('Il7r','Ccr7')),
data.frame(type='CD8',gene=c('Cd8a')),
data.frame(type='CD4', gene = c("Cd4")),
data.frame(type='Cycle',gene=c('Mki67','Top2a','Birc5'))
)
summarize(LCMV1)
cluster_names <- get.cluster.names(LCMV1, types, min.fold = 1.0, max.Qval = 0.01)
LCMV1 <- set.cluster.names(LCMV1, names = cluster_names)
LCMV2 <- setup_LCMV_example("LCMV2")
LCMV2 <- get.variable.genes(LCMV2, min.mean = 0.1, min.frac.cells = 0,
min.dispersion.scaled = 0.1)
LCMV2 <- PCA(LCMV2)
LCMV2 <- cluster.analysis(LCMV2)
summarize(LCMV2)
cluster_names <- get.cluster.names(LCMV2, types, min.fold = 1.0, max.Qval = 0.01)
LCMV2 <- set.cluster.names(LCMV2, names = cluster_names)
pooled_env <- setup_pooled_env()
pooled_env <- read.preclustered.datasets(pooled_env)
pooled_env <- PCA(pooled_env, clear.previously.calculated.clustering = F)
summarize(pooled_env, contrast = "datasets")
cluster.similarity <- assess.cluster.similarity(pooled_env)

robustSingleCell documentation built on May 2, 2019, 2:11 p.m.