Typically, after FastQC, Trimming, and STAR alignment have been performed.
Need to make a raw counts table from the STAR output, explore data, differential analysis.
-
Utilize FQC_Trim_align scripts to perform via SLURM: Performs:
- FastQC on raw files
- Trimmomatic
- FastQC on trimmed files
- alignment and gene counts via STAR and ensembl Process:
- Make metadata excel file with columns Sample_Name (1st column). Group, and fastq - technically don't need this until after alignment, but good to plan it now
- Locate the 3 scripts from FQC_Trim_Align to make args_file and execute SLURM (copy to work_dir/scripts)
- Edit make_rnaseq_args_file.sh script (PE or SE) to find proper fastq.gz file names
- Prepare 3 input variables: 1. path to raw fastq files 2. path to folder containing STAR indices for reference genome 3. path to gtf file for reference genome annotation
- execute submit_script_rnaseq.sh with the above 3 input variables in order on exacloud
-
Collect Stats
- trimming stats, STAR stats, fastqc stats
- scripts in RNAseq repository, collect_stats folder
- TODO: make script to collect fastqc stats
-
Utilize functions in bulkRNA_ide_source_functions.R to:
- create raw counts table from STAR output
- perform initial data exploration on the count data
-
Utilize functions to perform differential expression analysis
- with DESeq2 or edgeR
1. metadata excel file
- first column must be desired sample name (column Sample_Name)
- must contain column fastq which is the fastq file prefix
- this is the first part of the fastq file name
- (the part preceding the general suffix ".star.ReadsPerGene.out.tab" of the STAR output
- currently, uses column Group to define sample group membership (colors of PCA, groups to compare, etc)
- rows (samples) are in a desired order that makes sense (with replicates together)
- the sample columns of raw counts table will be in this same order
2. raw counts table
- made with compile_readcounts function and passing metadat excel file to get sample names
3. optional gene_info file
- provides matching gene names and gene descriptions for GeneIDs
Projects for reference:
- ECP55 (non-ensembl genome and annotation, make gene info file optional?)
- agarwal bulk RNAseq (clean up functions to be run within R, not command line)
- new collection method of FastQC, Trimming, STAR stats?
TODO:
- in max counts barplot function, how to correctly output txt file with different gene_info files - use descriptions?
- in ide.R script, make sure export of filtered counts is in good format (include GeneID as first column?)
- export cpm and logcpm counts in ide.R script? Currently exported in differential analysis