Using salmon to processes P. aeruginosa RNA-seq data

1. Create a conda environment with salmon installed called salmon2

   • conda create salmon2
   • conda activate salmon2
   • conda install -c bioconda salmon
   • (https://anaconda.org/bioconda/salmon)

2. Create transcriptome index These are reference genome specific and can be made in decoy-unaware or decoy-aware modes examples: salmon_ti.script, salmon_ti_decoy.script

3. Load fastq files to directories accessible on discovery

   • keep all fastq files that correspond to a sample (eg paired) in sub-directories
   • dartfs storage is accessible by discovery but larger than user home dirs but is usually specific to a lab
   • global scratch can also be used (files last 30 days)
   • examples: /dartfs-hpc/rc/lab/H/HoganD or /dartfs-hpc/scratch/

4. Map samples to t-index

   • Use the salmon quant command for fast mapping parameters used in this project were chosen for P. aeruginosa and may be bacteria-specific.
   • Run samples in parallel via job arrays.
   • examples: salmon_metals_spu.scipt, salmon_metals_spu_decoy.script

5. Collect salmon output (quant.fs files) into count table

   • the salmon calls save all outputs to files named quant.fs in different directories
   • almon also uses cDNA gene identifiers
   • collect them all as columns in a table with column names as sample names and rownames as gene names
   • example: spu_collect.script