generate_nmf_modules: Generate NMF Modules

In MollahLab/gen3DNet: An R Package for Generating 3D Network Models

Generate NMF Modules

Description

This function uses NMF to generate functional modules representing interactions between the entities in the rows and columns of the dataframe. This has two components: Each entity is assigned to a cluster "Column signatures" are generated for each row object. For example, if rows are histones and columns are drugs, this function will generate "histone signatures" that express Because NMF relies on knowing the number of clusters (k) in advance, this function also determines this automatically. Currently, the default option uses the number of clusters that produces the lowest KL index with Ward clustering. The package depends on NbClust to implement this feature. However, the user can override this option with a custom function.

Usage

generate_nmf_modules(
  left_data,
  nmf_nrun,
  k_range,
  k_picker = gen3DNet::max_ward_kl,
  seed,
  verbose = FALSE
)

Arguments

left_data	This is the data that will be clustered using NMF. Signatures will be generated for the object type that is represented by rows.
nmf_nrun	The number of r
k_range	Either a) A consecutive range of possible k values, b) a single k value, or c) nothing, meaning that generate_nmf_modules will assume the range is from 2 to the number of rows.
k_picker	Any function that takes a dataframe and a consecutive range of potential k values
seed	The seed to use with NMF
verbose	Whether to print output.

MollahLab/gen3DNet documentation built on Nov. 18, 2024, 5:08 a.m.