====Salmon====

Salmon is a tool for transcript quantification from RNA-seq data. To use Salmon all you need is a ''FASTA'' file containing your reference transcript, and a file or files containing your ''FASTA''/''FASTAQ'' reads. There are other files that can be used to run Salmon, documentation on this can be found [[http://salmon.readthedocs.io/en/latest/salmon.html|here]].

====Enviornment Setup====

Edit your ''$HOME/.bashrc'' file to include the load command for Salmon. You will have to log back in for this to take effect.

<code>
module load salmon/0.8.2
</code>

In your ''$HOME'' directory create a directory ''SALMON-JOBS'' where we will run the example job below, and copy the sample data provided from Salmon into this directory. Go into the ''$HOME/SALMON-JOBS'' directory to unzip and open the tar file which will open the ''sample_data'' directory.

<code>
razor-l3:jokinsey:~$ mkdir SALMON-JOBS
razor-l3:jokinsey:~$ cp /share/apps/bioinformatics/salmon/Salmon-0.8.2_linux_x86_64/sample_data.tgz SALMON-JOBS/
razor-l3:jokinsey:~$ cd SALMON-JOBS/
razor-l3:jokinsey:~/SALMON-JOBS$ tar -xzf sample_data.tgz
razor-l3:jokinsey:~/SALMON-JOBS$ cd sample_data
</code>

====Example Job====

In the ''$HOME/SALMON-JOBS/sample_data'' directory create a ''PBS'' script to submit the job. The script should look like the on below.

<code>
#!/bin/bash
#PBS -N salmon
#PBS -q tiny12core
#PBS -j oe
#PBS -o salmon.$PBS_JOBID
#PBS -l nodes=1:ppn=12
#PBS -l walltime=1:00:00
#PBS -M jokinsey@email.uark.edu

cd $PBS_O_WORKDIR
cp {reads_1.fastq,reads_2.fastq,transcripts.fasta} /scratch/$PBS_JOBID
cd /scratch/$PBS_JOBID

salmon index -t transcripts.fasta -i transcripts_index
salmon quant -i transcripts_index -l IU -1 reads_1.fastq -2 reads_2.fastq -o out

cp -r out $PBS_O_WORKDIR/out.$PBS_JOBID
</code>

The script above copies the files we will be working with the the directory we will be working in ''/scratch/$PBS_JOBID'', after that it creates an index from the ''transcripts.fasta'' file. Then we run salmon with the index we just build and the reads from the sample date. The information will be in a directory called ''out''. Lastly we copy the directory to the directory we ran the job from, and append the ''$PBS_JOBID'' to the new output directory so that we make a unique directory. This ensures we don't copy into an already created directory if you wanted to run this script more than once.

All that left is to submit the job.

<code>
razor-l3:jokinsey:~/SALMON-JOBS/sample_data$ qsub salmon.pbs
</code>

In the ''$HOME/SALMON-JOBS/sample_data/out.$PBS_JOBID/'' directory which will have the output from the job the most important file is the ''quant.sf'' file. More information on the output can be found [[http://salmon.readthedocs.io/en/latest/file_formats.html|here]].