A pipeline to analyse transcriptomic time-course data

An automatically compiled version of the manuscript's Rmd with Travis-CI can be found here: https://nellev.github.io/2019timecourse-rnaseq-pipeline/

1. Install the dependencies and the package.

   • Install all the dependances at once using the following command: Rscript scripts/install.R
   • Install timecoursedata via the install_github commadd install_github("NelleV/timecoursedata", dependencies=FALSE)
   • Install moanin via the install_github command install_github("NelleV/moanin", dependencies=FALSE))

2. Compile the manuscript:

   cd scripts make

Or, directly in R:

    library(rmarkdown)
    render("manuscript.Rmd", output_file="reports/manuscript.pdf")

3. Other elements: Note that some elements, such as the bootstrapped k-means, the consensus plots, and the pathway and GO term enrichment are pre-computed. To run everything from scratch:

   • Run the clustering for k between 2 and 20, and bootstrapped random seed between 1 and 30:

     • On a cluster, using the script scripts/cluster_scripts/stability_kmeans.sh
     • On a local machine, with:

   for SEED in {1..30};
   do
       for N_CLUSTERS in {2..20};
       do
   	Rscript run_clustering.R ${N_CLUSTERS} ${SEED}
       done;
   done;

   • Merge all results in a single file using the command:

   Rscript concatenate_clustering_stability_results.R

   • Compute the consensus plots:

   Rscript run_consensus_plots.R

   • Compute the pathway enrichment analysis for all clusters

   Rscript run_pathways.R

   • Compute the GO term enrichment analysis for all clusters

   Rscript run_go_terms.R

SUPPLEMENTARY: PThe normalization of the micro-array data

You can find an Rmd file containing detailed steps for the normalization in scripts/miscellaneous/normalization.Rmd