The python program 'pone.0142147.s003.py' accompanies the paper G Mahajan and SC Mande, "From System-Wide Differential Gene Expression to Perturbed Regulatory Factors: A Combinatorial Approach", PLoS ONE (2015), https://doi.org/10.1371/journal.pone.0142147.
The code takes as input a tab-delimited file with gene names/IDs (one name per line) along with binarized information on significant differential expression between two conditions/experiments (0 or 1), and a second tab-delimited file specifying the transcriptional regulatory network (one transcription factor/TF -> gene target per line). The gene differential expression data is overlaid on the precompiled reglatory network, and TFs significantly associated with the differentially expressed gene set are identified through application of Fisher's exact test in an individual manner, as well as the proposed approach to identifying the top-scoring subset of TFs. The latter is estimated through application of greedy search (methods A-C) on the exponentially large search space comprising all possible TF combinations (all sizes) among the precompiled regulatory network.