STAR alignment speed?

Question

[a_bis]

Hello, I'm currently running some RNA-seq experiment alignments on a server with STAR. After about two hours, looking at the head of the Log.progress.out file, I am getting a result like this:

My total read count for the sample is ~80,000,000 reads. Does this mean that judging by the progress in the 'Read number' column my alignment is going to take over a week? It feels like this is abnormally long. When I submitted the bash script to the queue, the script started as follows:

#!/bin/bash
#SBATCH --job-name=star
#SBATCH --ntasks=1
#SBATCH --cpus-per-task=16
#SBATCH --time=0-64:00:00
#SBATCH --output=outfile.%j
#SBATCH --error=errfile.%j

module load STAR/2.7.4

STAR --genomeDir mm39starindex/mm39starindex/index --readFilesIn WTCHG_894903_72185194_1.fastq.gz WTCHG_894903_72185194_2.fastq.gz --readFilesCommand zcat --outSAMtype BAM SortedByCoordinate --quantMode GeneCounts --outFileNamePrefix alignments/WTCHG_894903_72185194

Should I have specified a longer time or more CPUs? Any advice will be really appreciated! Many thanks.

[igor]

Less than 10% unique mapped is very low. You may want to look into that as well.

[a_bis]

Thank you for the advice. I'm very new to RNA-seq read alignment -- in your experience is there a particular thing I could do to improve alignment rates? Thank you!

[igor]

Looks like you already posted a follow-up question: Low mapping frequency on STAR

[a_bis]

Hi, yes, my question subsequently changed so I thought it would be best to rephrase it. Thanks again for your advice!

Answer 1

[rpolicastro]

You need to specify the number of threads STAR uses with --runThreadN 16.

[a_bis]

Thanks, that's super helpful!