==================
Performs an integrative Causal Regulatory Network Analysis (CRNA) of system level measurements (He at al. 2024).
==================
Causal Regulatory Network Analysis (CRNA) explains observed gene expression changes by a causal pattern under a given experiment. Recent approaches use a pre-established Knowledge Base to extract the regulatory network that is consistent with Differentially Expressed Genes (DEGs) in a given study. However, these methods consider all information in their Knowledge Base, regardless of the experimental context, leading to inconsistencies in regulatory network predictions. An efficient and effective method is needed to deliver regulatory annotations and activity analyses from system level expression measurements of high-throughput data such as generated by Genomics, Proteomics, or Metabolomics studies.
We introduce a novel Causal Regulatory Network framework to predict a causal pattern of events for the provided systems under study and infer Genes/Proteins activities. Given a directed Knowledge Base network of Gene-Gene or Protein-Protein interactions, we provide a tunable parametric graphical model, based on network-searching algorithms, to build a causal regulatory network and infer plausible scores/states of Genes/Proteins activity. We assess the predicted upstream regulators at the Gene/Protein level and regulatory paths at the sample and phenotype level. Finally, we report the regulatory network with predicted activity score/state for each Gene/Protein assessed in the study. Simulation results also show a significant improvement in prediction accuracy of regulatory nodes in the final network of Genes/Proteins.
Our framework addresses several issues prevalent in Causal Regulatory Network analysis methods. First, we provide the necessary tuning technique of model parameters by cross-validation to avoid the common risks of over and under fitting in graphical models. Second, we improve on the inconsistencies in activity states that are often observed in these models. Moreover, our method can predict activity scores/states for Genes/Proteins and paths for an individual sample or multiple measurements of the same phenotype, which can also be used to assess the quality of samples, providing flexibility for downstream analyses.
==================
- GitHub (main) or GitHub (mirror site).
- Supplementary information: Supplementary data and methods are available on this site in the data folder.
==================
Experimental designs with small sample size per group might lead to problems due to the current implementation of the bootstrapping resampling. We recommend using at least 5 samples per group. This limitation will be addressed in a future release of the software soon.
==================
This branch (master) is the default one, that hosts the current development release (version 1.0.0).
==================
CRNA is at this time open source / free software, licensed under the GNU General Public License version 3 (GPLv3), sponsored by the Free Software Foundation. To view a copy of this license, visit GNU Free Documentation License.
==================
==================
CRNA (>= 1.0.0) requires R-3.5.0 (2018-04-23). It was built and tested under R version 4.3.1 (2023-06-16) and GitHub Actions CI.
The installation has been tested on Windows, Linux, OSX platforms.
See GitHub Actions CI build result:
To satisfiy required packages, before installation try the following:
setdiff(c("Rcpp", "parallel", "limma", "Matrix", "abind", "igraph", "glmnet", "testit", "magrittr", "stats", "utils", "BH"), rownames(installed.packages()))
You will need to install these packages before CRNA.
==================
- In the current state of
CRNA, please download the tarball including the source code (from the Releases section on the right side) and install the package using the following command:
R CMD INSTALL CRNA_1.0.0.tar.gz
- For recompilation and building after modifications to the source code, you can use the following command in the package directory to generate a tarball which can be installed again using the above command:
R CMD build .
==================
- To load the CRNA library in an R session and start using it:
library("CRNA")
- Check details of new features, changes, and bug fixes with the following R command:
CRNA.news()
- Check on how to cite the package with the R command:
citation("CRNA")
Refer to quickstart folder for an example on how to run CRNA on real RNA-seq data.
===================
===================
Authors:
- Jean-Eudes Dazard, Ph.D. jean-eudes.dazard@case.edu
- Dongze He, Ph.D., dhe17@umd.edu
- Furkan Eris, furkan.eris@sabanciuniv.edu
- Durmaz Arda, axd497@case.edu
- Gurkan Bebek, Ph.D. gurkan.bebek@case.edu
Maintainers:
- Furkan Eris, furkan.eris@sabanciuniv.edu
- Jean-Eudes Dazard, Ph.D. jean-eudes.dazard@case.edu
Funding/Provision/Help:
- This work made use of the High Performance Computing Resource in the Core Facility for Advanced Research Computing at Case Western Reserve University.
- This project was partially funded by the National Institutes of Health NIH - National Cancer Institute (P30-CA043703).
===================
- He D., Dazard J-E., Eris F., Durmaz A., and Bebek G. Integrative Network Analysis Discovers Causal Regulatory Network of System Level Measurements. submitted (2024), -(-):--.