run_ICLite: run_ICLite: Wrapper function for ICLite package

In camiolomj/ICLite: Immune Cell Linkage Through Exploratory matrices

Description

Runs the ICLite algorithm based on input parameters provided by user. Creates output graphical files based on gene module and cell interactions, CSV files of gene modules and global environment object "gene_module_lists" that may be used for downstream Gene Ontology analysis. Example data can be loaded using load_IMSA_data()

Finding the optimal solution for a data set may require running multiple iterations of ICLite with varying input parameters. Running separate tests for correlation between genes in a data set is recommended to determine minimum values for input rho exclusion. For general practice, we recommend using no lower than 0.4. The number of assumed gene clusters should be considered in relation to the total size of the transcriptional data set. Though ICLite does penalize for over-clustering, it will only consider solutions from the input vector. Therefore, initial runs may benefit from a broad array of values that may be narrowed on successive iterations.

Users should include a vector of at least 2 input parameters for minimum connectivity, rho cutoff and number of clusters.

Usage

run_ICLite(
  gene_expression_data,
  immune_cell_logratios,
  input_connectivities,
  input_rho,
  number_of_clusters
)

Arguments

gene_expression_data	A matrix of normalized gene expression data where columns represent individuals and rows represent features (e.g. genes)
immune_cell_logratios	A matrix of corresponding cell log ratios where rows represent individuals and columns represent features (e.g. cells). Transformation of percentage values using the "compositions" package is recommended prior to use
input_connectivities	A vector of minimum connectivity values. Higher cutoffs will result in smaller gene modules
input_rho	A vector of rho exclusion values ranging from 0.3 to 0.9. Gene correlations below this value are converted to 0 in binary space while those above are converted to 1. Higher cutoffs will result in smaller gene modules
number_of_clusters	A vector of assumed number of clusters to be used for blockclustering

Value

Module to gene connectivity solution and gene module memberships to working directory and global environment

camiolomj/ICLite documentation built on July 28, 2021, 6:33 a.m.