summarize_clusters: summarize_clusters

In clatworthylab/bowerbird: A single-cell package for Functional Annotation + Gene Module Summarization.

Description

Summarize the terms of the cluster using pagerank algorithm

Usage

## S3 method for class 'BOWER'
summarize_clusters(
  bower,
  cluster = NULL,
  pattern = NULL,
  sep = NULL,
  ncpus = NULL,
  disconnect_graph = FALSE,
  ...
)

## S3 method for class 'igraph'
summarize_clusters(
  graph,
  cluster = NULL,
  pattern = NULL,
  sep = NULL,
  ncpus = NULL,
  disconnect_graph = FALSE,
  ...
)

Arguments

cluster	vector of cluster labels for each geneset.
pattern	search pattern to remove from the terms. Unless specified, will default to built-in pattern.
sep	separator used/found in gene set names to be changed to blank spaces. Default value is underscore ('_').
ncpus	number of cores used for parallelizing reconstruction.
disconnect_graph	return a graph connecting only nodes in a cluster.
...	passed to textrank::textrank_sentences.
graph	geneset overlap graph.

Details

Given a list of text, it creates a sparse matrix consisting of tf-idf score for tokens from the text. See https://github.com/saraswatmks/superml/blob/master/R/TfidfVectorizer.R. A k shortest-nearest neighbor graph is then computed using the overlap of of the terms.

Value

Returns a matrix of tf-idf score of tokens.

Examples

gmt_file <- system.file("extdata", "h.all.v7.4.symbols.gmt", package = "bowerbird")
bwr <- bower(gmt_file)
bwr <- snn_graph(bwr)
bwr <- find_clusters(bwr)
bwr <- summarize_clusters(bwr, ncpus = 1)
bwr

clatworthylab/bowerbird documentation built on Dec. 19, 2021, 5:15 p.m.