Date created: | Last Updated:

Category: Project

Description: This analysis is performed using Trimmomatic, Kallisto, Sleuth, and Gage. The scripts are designed to execute this analysis on a supercomputer over ssh. Files beginning with "run" contain code for executing the programs on the supercomputer. Files ending in "sbatch.sh" submit jobs to the supercomputer. Scripts should be run in the following order: 1. trimmomatic_sbatch.sh 2. kallisto_sbatch.sh 3. sleuth_array.sh All other scripts are called within the scripts listed above. Description of Scripts: trimmomatic_sbatch.sh submits a job to the supercomputer to run trimmomatic using run_trimmomatic.sh. Trimmomatic removes adapters and bad quality reads from the RNA-seq data. The output from trimmomatic is compiled in a folder called "trimmomatic_output." kallisto_sbatch.sh submits a job to the supercomputer to run kallisto using run_kallisto.sh. Kallisto performs a pseudoalignment of the RNA-seq reads and quantifies transcript abundances. The output from kallisto is compiled in a folder called "kallisto_output." sleuth_array.sh executes multiple scripts that run sleuth and gage and perform a variety of transformations on the data that are described below. Before running sleuth, metadata files must be written for each comparison. These files should list the samples used in each comparison and which group each of the samples belongs to (control or mutant). Data must be sorted into comparison folders. If metadata files are written properly, this can be completed using copyData.R. After the data is sorted, sleuth can be run using sleuth_script.R. In our analysis, we were interested in fold change values, so we used a function to transform b values to fold change values. create_abundance_table.R structures the data in a more useable format. convert_to_symbols.py converts Ensembl gene IDs to symbols. gage_script.R performs a gene ontology analysis. filter_transcripts.py identifies the type for each transcript and adds that information to the data. edit_file.R renames the "b" column to "fold_change." These values were converted into fold change values in sleuth_script.R. convertGageToExcel.R saves the gage output file as an excel file. visualize.sh creates heatmaps and volcano plots for each of the comparisons.