Inference of bacterial small non-coding RNA networks using the Inferelator
This repository contains all the input files, Inferelator code and the R scripts used for generating the results reported in the revised version of the paper:
Arrieta-Ortiz ML, Hafemeister C, Shuster B, Baliga NS, Bonneau R, Eichenberger P. 2019. Inference of bacterial small RNA regulatory networks and integration with transcription factor driven regulatory networks. bioRxiv. doi: https://doi.org/10.1101/657478
Below, we briefly describe the contents of this repository.
1. Input_files: it contains the files required to run the Inferelator in each the four analyzed bacterial species. Files for inferring the Escherichia coli network that includes eight sRNAs are in the Ecoli_8sRNAs folder. Similarly, files for inferring the FsrA, PrrF and RsaE regulons for Bacillus subtilis, Pseudomonas aeruginosa and Staphylococcus aureus are in the corresponding folders. The input folders for the instances of shuffled E. coli sRNA priors, CopraRNA-derived sRNA priors (for RyhB, GcvB and Spot 42) and noisy E. coli sRNA priors (with false sRNA priors added) are included. The ratio of true: false sRNA priors (1:1, 1:2, 1:5) is indicated by the ni term in last part of the noisy_sRNA_priors_1_*.csv files. Although we did not infer sRNA regulons for S. enterica, the files required to infer a transcriptional regulatory network for this species (sRNA priors are not included) are available.
Each folder contains an expression matrix (*.RData), TFs and sRNA priors (also refer to as gold standard-gs) and a .csv file with the list of all potential regulators to be considered (TFs and sRNAs). The values in the prior matrix are in the {0, 1, -1} set, where 1 and -1 indicate activation and repression, respectively.
2. Inferelator_code: this folder contains the code used to infer all the reported TF-controlled and sRNA-controlled networks. For more information about how to run the Inferelator, please check the README file in this folder.
This code was previously used for generating the improved transcriptional regulatory network model of B. subtilis reported in: Arrieta-Ortiz ML, Hafemeister C, et al. 2015. An experimentally supported model of the Bacillus subtilis global transcriptional regulatory network. Mol. Syst. Biol. 11.
This code is also available at: https://github.com/ChristophH/Inferelator
3. Inferelator_output_files: it contains all the networks inferred with the Inferelator (but the ones with noisy E. coli priors). Each output folder contains three files: regression coefficients (betas_*.RData files), confidence scores (combined_conf*.RData files) and the parameters and input file (params_and_input.RData).
4. Miscellaneous_scripts: it contains multiple R scripts that can be used to analyze the output of the Inferelator.
Finally, this repository includes an R notebook with the code for post-inference analyses.