CrossBatchGrouping: Cross Batch ICA Source Component Grouping

In WWXkenmo/ICAnet: What the Package Does (One Line, Title Case)

Cross Batch ICA Source Component Grouping

Description

Group the expression programs (source components) generated from different batch of scRNA-seq dataset through PAM clustering.

Usage

CrossBatchGrouping(
  ica.pooling,
  k.max = (ncol(ica.pooling) - 1),
  plot = TRUE,
  cor = "pearson",
  W.top = 2.5,
  filtering = TRUE,
  threshold = 30,
  Unique.Preservation = TRUE
)

Arguments

ica.pooling	A matrix object that including all ICA source components generate from ICAcomputing,
k.max	the maximum clusters number of ica-components (default: the number of ica-components - 1)
plot	whether plot cross batch ica source component correlation map
cor	the correlation method which determine the correlation between different pair of source components
W.top	the threshold to determine the activated genes, the genes which has absolute attributes value large than threshold*standard derivation from mean are the activated genes (default: 2.5)
filtering	whether filter the components which their 'activated gene' number lower than a specific values
threshold	the threshold number to determine which component need to be filtered (default: 30)
Unique.Preservation	whether to preserve the cluster with only one component included (default: TRUE)

Value

a list object which containing the filtered ica-components matrix.

WWXkenmo/ICAnet documentation built on April 11, 2022, 5:44 a.m.