gnet: Run GNET2

Description Usage Arguments Value Examples

View source: R/build_module.R

Description

Build regulation modules by iteratively perform regulator assigning and Gene assigning, until the assignment of genes did not change, or max number of iterations reached.

Usage

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
gnet(
  input,
  reg_names,
  init_method = "boosting",
  init_group_num = 4,
  max_depth = 3,
  cor_cutoff = 0.9,
  min_divide_size = 3,
  min_group_size = 2,
  max_iter = 5,
  heuristic = TRUE,
  max_group = 0,
  force_split = 0.5,
  nthread = 4
)

Arguments

input

A SummarizedExperiment object, or a p by n matrix of expression data of p genes and n samples, for example log2 RPKM from RNA-Seq.

reg_names

A list of potential upstream regulators names, for example a list of known transcription factors.

init_method

Cluster initialization, can be "boosting" or "kmeans", default is using "boosting".

init_group_num

Initial number of function clusters used by the algorithm.

max_depth

max_depth Maximum depth of the tree.

cor_cutoff

Cutoff for within group Pearson correlation coefficient, if all data belong to a node have average correlation greater or equal to this, the node would not split anymore.

min_divide_size

Minimum number of data belong to a node allowed for further split of the node.

min_group_size

Minimum number of genes allowed in a group.

max_iter

Maxumum number of iterations allowed if not converged.

heuristic

If the splites of the regression tree is determined by k-means heuristicly.

max_group

Max number of group allowed for the first clustering step, default equals init_group_num and is set to 0.

force_split

Force split the largest gene group into smaller groups by kmeans. Default is 0.5(Split if it contains more than half target genes)

nthread

Number of threads to run GBDT based clustering

Value

A list of expression data of genes, expression data of regulators, within group score, table of tree structure and final assigned group of each gene.

Examples

1
2
3
4
5
6
7
8
9
set.seed(1)
init_group_num = 8
init_method = 'boosting'
exp_data <- matrix(rnorm(50*10),50,10)
reg_names <- paste0('TF',1:5)
rownames(exp_data) <- c(reg_names,paste0('gene',1:(nrow(exp_data)-length(reg_names))))
colnames(exp_data) <- paste0('condition_',1:ncol(exp_data))
se <- SummarizedExperiment::SummarizedExperiment(assays=list(counts=exp_data))
gnet_result <- gnet(se,reg_names,init_method,init_group_num)

GNET2 documentation built on Nov. 8, 2020, 8:03 p.m.