Regularized Gradient Boosting Machine for inferring GRN
Get indices of experiments where knockout or knockdown happened
Identifies the optimal value of k i.e. top k Tfs for each target gene
Re-iterate through the core GBM model building with optimal set of Tfs for each target gene
Perform the regularized GBM modelling once the initial GRN is inferred
Regulate the size of the regulon for each TF
Get the indices of all the TFs from the data
Generates a matrix S2 of size Ntfs x Ntargets using the null-mutant zscore algorithm Prill, Robert J., et al
Column normalize the obtained adjacency matrix
null_model_refinement_step
Perform the null model refinement step
Train the regression stump
Test the regression model
Generate filepaths to maintain adjacency matrices and images
transform_importance_to_weights
Log transforms the edge-weights in the inferred GRN
Convert adjacency matrix to a list of edges
Get the indices of recitifed list of Tfs for individual target gene
Calculate Gene Regulatory Network from Expression data using either LS-TreeBoost or LAD-TreeBoost
Test rgbm predictor
Apply row-wise deviation on the inferred GRN
consider_previous_information
Remember the intermediate inferred GRN while generating the final inferred GRN
Train GBM predictor
Perform either LS-Boost or LAD-Boost (GBM
) on expression matrix E followed by the null_model_refinement_step
Add row and column names to the adjacency matrix A
Train RGBM predictor
Test GBM predictor