This repository contains code and data associated with the PSB 2016 conference paper entitled: Exploring the Reproducibility of Probabilistic Causal Molecular Network Models
##Data:
https://github.com/divara01/PSB2017_ReproducibilityOfBNs/tree/results/*MappedCliqueList.RData - contains a list format of all the cliques that were detected using COS.
https://github.com/divara01/PSB2017_ReproducibilityOfBNs/tree/results/*edges_all_RIMBANet.RData - a data frame object containing all the edges for all the different BNs (for each dataset). Format is node1, node2, posterior probability, replication, subsampling type.
https://github.com/divara01/PSB2017_ReproducibilityOfBNs/tree/results/*edges_all_RIMBANet_cliqueAdded.RData - a data frame object containing all the edges for each dataset, but appended to it is a column if both nodes in an edge are present in a clique.
https://github.com/divara01/PSB2017_ReproducibilityOfBNs/tree/results/*edges_all_RIMBANet_cliqueAdded_KDAdded.RData - a dataframe where appended is if node1 is a Key Driver.
https://github.com/divara01/PSB2017_ReproducibilityOfBNs/tree/data/*BC - The continuous data for the input of running BN. Format is Genes as rows, no column names, but rows are labeled.
https://github.com/divara01/PSB2017_ReproducibilityOfBNs/tree/data/*BD - The discretized data, where the three states are 0,1,and 2. Same file format as the continuous data.
https://github.com/divara01/PSB2017_ReproducibilityOfBNs/tree/data/*eqtl.txt - the list of genes which have an eQTL associated with them
https://github.com/divara01/PSB2017_ReproducibilityOfBNs/tree/data/*citPriors.txt - priors for the STARNET data that we detected using the Causal Inference Test. (See PMID: 27540175 for methods)
https://github.com/divara01/PSB2017_ReproducibilityOfBNs/tree/data/*EdgeFile - these are the output for RIMBANet of the consensus digraph where the BN is a directed acyclic graph (DAG). Edges here don't have weights.
https://github.com/divara01/PSB2017_ReproducibilityOfBNs/tree/data/*labeledEdges - output from RIMBANet where the label is the number of times out of 1,000 reconstructions that the edge is present in.
https://github.com/divara01/PSB2017_ReproducibilityOfBNs/tree/data/Simulation_true_network.txt - the actual simulation true, real network. ##Code: ####To generate all the plots and tables for the PSB paper - https://github.com/divara01/PSB2017_ReproducibilityOfBNs/blob/master/PSB_RMarkdownGIT.Rmd
####To run the clique detection - We used COS https://sourceforge.net/projects/cosparallel/ with the standard parameter settings.
####To Download RIMBANet - http://research.mssm.edu/integrative-network-biology/RIMBANET/RIMBANET_overview.html
####Scripts used for running RIMBANet - #####Driver script -- https://github.com/divara01/PSB2017_ReproducibilityOfBNs/blob/master/BN_driver_extra.sh #####Readme for the Driver script -- https://github.com/divara01/PSB2017_ReproducibilityOfBNs/blob/master/BN_extra.README #####Script called in the driver -- https://github.com/divara01/PSB2017_ReproducibilityOfBNs/blob/master/process_w_genetics_extra.pl #####Script called in the driver to submit jobs to LSF queue system -- https://github.com/divara01/PSB2017_ReproducibilityOfBNs/blob/master/run_w_genetics_queue.pl #####All other scripts will be downloaded when RIMBANet is downloaded.
####To run RIMBANet for each of the subsampling networks - https://github.com/divara01/PSB2017_ReproducibilityOfBNs/blob/master/PSB_Run_BN.sh