I’m processing Salmon RNAseq data that match my Salmon methyltion data so I can perform an integrated analysis of these.

I’m using the nf-core RNAseq pipeline.

On Mox:

I started installing it on Mox and got to the initial pipeline test, but this failed because it couldn’t allocated enough memory (which I suspect is because I was running it from the command prompt rather than as a script). Unfortunately the nodes were all being used, so I decided to try running it on Roadrunner.

On Roadrunner:

This was generally a smooth installation process. I had to install docker because I couldn’t get it to work with conda. I had one hangup with docker and it was a permissions thing that I was able to resolve following a github issue (noted in log file from installation). The RNAseq pipeline test completed without errors and the RNAseq pipeline is currently running on the raw data on Roadrunner

Log file from installation here:


Running RNAseq pipeline on Salmon data

  • had to adjust the max memory allocated to processes because this computer only has 48GB available
    • I did this by setting the –max-memory parameter to ‘48.GB’
srlab@roadrunner:~/Documents/Shelly/Ssalar nextflow run nf-core/rnaseq -profile docker \
--max_memory '45.GB'
--reads '*_{1,2}.fastq.gz' \
--aligner hisat2 \
--fasta GCF_000233375.1_ICSASG_v2/GCF_000233375.1_ICSASG_v2_genomic.fa \
--gtf GCF_000233375.1_ICSASG_v2/GCF_000233375.1_ICSASG_v2_genomic.gtf \
--gff GCF_000233375.1_ICSASG_v2/GCF_000233375.1_ICSASG_v2_genomic.gff \
--bed12 GCF_000233375.1_ICSASG_v2/GCF_000233375.1_ICSASG_v2_MASTER.bed \
--saveReference \
--saveTrimmed \
--saveUnaligned \

~12 hours to complete using 45gb ram