Skewer - MattHuff/scRNASeq_011224 GitHub Wiki

These reads had more adapter contamination and lower quality scores than the previous run of scRNASeq, so I chose to perform a trimming step before running Cell Ranger. I used the program Skewer to trim these reads as opposed to Trimmomatic from my ATACSeq analysis to leave more of the reads in tact after trimming.

Directory Set-Up

In your main analysis directory, create a new sub-directory for trimming:

mkdir 2_trim
cd 2_trim

Installing Skewer

As Skewer is only available through GitHub, install it using git clone:

git clone https://github.com/relipmoc/skewer

Create the executable

cd skewer
make
cd ..

Link Trimmomatic Adapters

ln -s /zfs/musc3/huffmat/ATAC-Seq/analysis/2_trimmomatic/all.fa

1. Run Skewer

I ended up submitting each job in separate qsub submissions. The generally guideline of these script looks like this:

#!/bin/bash

#PBS -N 2_skewer_sample2
#PBS -l walltime=01:30:00
#PBS -l select=1:ncpus=16:mem=200gb
#PBS -j oe

source ~/.bashrc
cd $PBS_O_WORKDIR

/zfs/musc3/huffmat/Jordan_scRNASeq_011024/analysis/2_trim/skewer/skewer \
	-x /zfs/musc3/huffmat/Jordan_scRNASeq_011024/analysis/2_trim/all.fa \
	-t 16 \
	-l 30 \
	-Q 30 \
	/zfs/musc3/huffmat/Jordan_scRNASeq_011024/raw_data/LVF006-Septum_S1_L001_R1_001.fastq.gz \
	/zfs/musc3/huffmat/Jordan_scRNASeq_011024/raw_data/LVF006-Septum_S1_L001_R2_001.fastq.gz \
	-o LVF006-Septum_S1_L001 \
	>& LVF006-Septum_S1_L001.trim_output

I attempted to run this on all four samples, but it went beyond available memory. In cases like this, it is best to run the job in separate qsub submissions. Just be sure to change the input options (between -Q and -o), and then change the output options to make sure it will not overwrite the previous runs.