Hi there,
I have a series of RNA-seq samples and, when I perform the alignment with STAR (version 2.7.9a), only a little percentage of reads is aligned...
FastQC shows a high number of overrepresented sequences related with rRNA (determined manually by BLAST) and a weird "Per base sequence content". I have two questions:
Is there any way to improve these mapping rates? Thanks in advance!
EDIT
The used protocol for rRNA depletion is the: Illumina Stranded Total RNA Prep with Ribo-Zero Plus...
and the STAR command used:
base_name=${filename%_R?_001.cutadapt.fastq.gz}
STAR --runThreadN 20 \
--outFilterMismatchNmax 3 \
--alignEndsType Local \
--outFilterMultimapNmax 10 \
--outMultimapperOrder Random \
--genomeDir $reference \
--readFilesIn ${base_name}_R1_001.cutadapt.fastq.gz ${base_name}_R2_001.cutadapt.fastq.gz \
--readFilesCommand zcat \
--outFileNamePrefix ${base_name}.star. \
--outSAMtype BAM SortedByCoordinate \
--outBAMsortingThreadN 10 \
--genomeLoad LoadAndKeep \
--outSAMunmapped Within \
--limitBAMsortRAM 40000000000
One thing to keep in mind is that the overrepresented sequences in a fastqc report are a subset of the total number of overrepresented sequences. The software looks at the first 100k reads, then any overrepresented sequences found in those first 100k are tracked throughout the rest of your fastq file. See here. In other words, summing up the "counts" in your fastqc report does not tell the full story, especially when a few of them are one nucleotide different from each other. Bottom line, your data likely have much more than 3 million rRNA reads.
That's absolutely true, thanks for remembering it!
So I guess if I try to align my samples versus a reference with ribosomal sequences I would "detect" if these unmapped reads correspond to rRNA, do you know where I can download a reference with rRNA regions, please?
I followed this guide: Creating ribosomal RNA reference sequence. So far it's worked well enough for me to monitor how my custom rRNA depletion protocol is working.
Log in to answer this question.
You may want to provide the STAR command you are running so we can double check the code, and more detail from the STAR results log so we can check if there is a possibility that rRNA contamination could be contributing to this. Also, how was rRNA depleted in your RNA-seq? Was it a ribosomal depletion kit, poly-dT priming, etc..
If the above has been checked and confirmed you can get a more accurate idea of rRNA contamination by filtering your reads with a program such as BBDuk.
The per base sequence content looks typical of RNA-seq, so I wouldn't worry about that part.
Thanks for the reply. I've just updated the post (good suggestion).
Also, try running the FastQScreen tool to check if it is actually rRNA contamination. ^_^