RGBM (version 1.0-7)

select_ideal_k: Identifies the optimal value of k i.e. top k Tfs for each target gene

Description

This function detects the optimal number of transcription factors which are regulating each target gene. This number is different for different target genes. It utilizes a heuristic to also detect the isolated targets which are not regulated by any transcription factor. To the detect the optimal number of Tfs for each target gene, it uses a notion similar to that used for optimization of the L-curve criterion for Tikonov regularization by evaluating the variable importance curve for each target gene.

Usage

select_ideal_k(experimentid, mink, filepath, imagepath, adjacency_matrix_path)

Arguments

experimentid

The id of the experiment being conducted. It takes natural numbers like 1,2,3 etc. By default it's 1.

mink

User specified threshold i.e. the minimum number of Tfs to be considered while optimizing the L-curve criterion. By default it's 0.

filepath

Path where some intermediate files will be written and provided by the function get_filepaths.

imagepath

Path where an image of the variable importance curves for first 16 target genes will be written and provided by the function get_filepaths.

adjacency_matrix_path

Path where an intermediate adjacency matrix will be written and provided by the function get_filepaths.

Value

Returns a vector where each element represents the optimal number of transcription factors for each target gene.