Hi rsafavi, How to prepare the rRNA fasta for metagenomics? Thanks.
• 0 views
•
link
written
by
Shicheng Guo
Filtering rRNA from RNAseq data
Hi everyone,
I am trying to see if my RNAseq fastq file has any rRNA contamination. The reason I think we might have contaminations is because my GC content plot from fastqc has many peaks, and it basically fails. I used sortmerna with their default euk rRNA database, and I found about 18 gig out of 25 gig reads to be rRNA. I am not sure if this is right, since my data is from mouse, but the database is for all euk. I would very much appreciate if someone can point me to where I can find mouse rRNA database? Would it be enough to use gencode annotation file to filter out rRNA annotations, and extract the corresponding fasta file of those annotation and use that as the database?
Thanks!
3 answers
RSeQC tool works with the bam file. It takes in an rRNA bed file and alignment file and it splits your reads into rRNA aligned reads, reads that did not align to rRNA, and qcfailed,unmapped reads. Still, at the level of fastq file, I think sortmerna is a good tool, but it requires rRNA fasta
You can find the mouse rDNA repeat unit sequence here. You can then use bbsplit.sh from BBMap suite to separate your reads into bins (A: Tool to separate human and mouse ran seq reads ). If you don't have genome sequence without the rDNA repeat then you could use bbmap.sh with rDNA reference and capture reads that don't match by using the outu= option.
There are some related suggestions in this previous thread: RNA-seq rRNA contamination
@igor: Have you seen this (Inviting Biostars moderators to join Biostars slack channel )?
Apologies for high-jacking this thread for a bit.
Log in to answer this question.
More posts like this
-
Bulk RNAseq Salmon index building which transcriptome to use
written by Orange •
-
Low mapping percentage
written by Sib •
-
BASH SortMeRNA--Fasta Must be followed by an index name
written by dcjordan128 •
-
When should one be concerned about the CG content in RNA-seq experiment?
written by dsastre •
-
SortMeRNA for multiple reads file
written by uvika85 •
-
Per sequence GC content failed in Human RNA-seq Data
written by tahseen.abbas •
-
Filtering rRNA contamination (indicated by GC content plots) from RNA-seq data
written by tryingmybest •
-
FastQC results with abnormal GC content and Overrepresented sequences in RIP-seq datasets
written by yoyostudents5775 •
-
FastQC results with abnormal GC content and Overrepresented sequences in RIP-seq datasets
written by bioinfo_wen •
-
sortmerna command not working
written by Shahzad •
Using blast you can search your data against to mouse rRNA sequences. for this you can use
remoteoption of blast withspeciesoption. You can specify mouse in the species option of blast.Thank you, I will try that
You can normally align the reads against genome + annotation (gencode), using STAR for example which can count the number of reads/feature. Then just check the annotated rRNA gene count percentage. You would need "good" samples as a control of what is normal/bad.
rRNA can cause to low mapping ratio of reads to genome.