Is a transcriptome mapping rate of 55% acceptable for my total RNA experiments? How to check mapping rate of genome alignment?

290 views
Skip to first unread message

William Wright

unread,
Jul 11, 2019, 12:55:49 PM7/11/19
to RSEM Users
Hi everyone,

I'm using Bowtie2 with RSEM. Illumina RNA-seq (total, not mRNA-enriched) paired-end . My transcriptome mapping rates are around 55% for all samples. I also need genome mapping, and the "--output-genome-bam" is very handy. However, I can't see the mapping rate for the genome. 

Additionally - I'm a bit worried because I used STAR on these samples in the past to map to the genome (Hg19 then) with >90% mapping. 
I'm not sure if this is acceptable or normal or if I should continue or stop the analysis. 

Furthermore - I tried Salmon and the transcriptome mapping was also around 55% - so I don't think it's RSEM specific. Here's my code :




$rsem -p 8 --bowtie2 --bowtie2-path $bowtie2 --bowtie2-sensitivity-level sensitive --paired-end --strandedness reverse --append-names --sort-bam-by-coordinate --output-genome-bam $indir/CLEAN_Sample1_R1.fastq.gz $indir/CLEAN_Sample1_R2.fastq.gz $reference $outdir/RSEM_hg38__Sample1.fastq.gz


Thanks!
Charlie

Peng

unread,
Jul 11, 2019, 1:10:49 PM7/11/19
to RSEM Users
Hi Charlie,

It seems to me that ~90% of your RNA-seq data mapped to genome (according to STAR) and ~55% of your RNA-seq data mapped to transcriptome (according to RSEM).  So there are ~35% of your RNA-seq data mapped to genomic regions without any transcript annotation.  Would this 35% something you expected?  Is it possible that this 35% could be largely ribosome RNA since your RNA-seq date is not mRNA-enriched? 

Best,
Peng

William Wright

unread,
Jul 11, 2019, 1:51:34 PM7/11/19
to rsem-...@googlegroups.com
Hi Peng,

thanks for the response. Indeed, it is possible that there is rRNA contamination, but I'm just not well-versed enough in transcriptome-mapping to know if this is what others get. How can I check the mapping rate of Bowtie2/RSEM's genome mapping step? 

Thanks!
Charlie

--
RSEM website: http://deweylab.biostat.wisc.edu/rsem/
---
You received this message because you are subscribed to the Google Groups "RSEM Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rsem-users+...@googlegroups.com.
To post to this group, send email to rsem-...@googlegroups.com.
Visit this group at https://groups.google.com/group/rsem-users.
To view this discussion on the web visit https://groups.google.com/d/msgid/rsem-users/2049a45e-8d0b-41db-9627-293196c5073f%40googlegroups.com.

Peng

unread,
Jul 11, 2019, 2:06:56 PM7/11/19
to RSEM Users
Hi Charlie,

I don't know how to check the genome mapping rate for Bowtie2/RSEM.  If you use STAR/RSEM, you can get the genome mapping rate from STAR's 'Log.final.out' file.  Hope this information is helpful.

Best,
Peng


On Thursday, July 11, 2019 at 12:51:34 PM UTC-5, William Wright wrote:
Hi Peng,

thanks for the response. Indeed, it is possible that there is rRNA contamination, but I'm just not well-versed enough in transcriptome-mapping to know if this is what others get. How can I check the mapping rate of Bowtie2/RSEM's genome mapping step? 

Thanks!
Charlie

On Thu, Jul 11, 2019 at 12:10 PM 'Peng' via RSEM Users <rsem-...@googlegroups.com> wrote:
Hi Charlie,

It seems to me that ~90% of your RNA-seq data mapped to genome (according to STAR) and ~55% of your RNA-seq data mapped to transcriptome (according to RSEM).  So there are ~35% of your RNA-seq data mapped to genomic regions without any transcript annotation.  Would this 35% something you expected?  Is it possible that this 35% could be largely ribosome RNA since your RNA-seq date is not mRNA-enriched? 

Best,
Peng


On Thursday, July 11, 2019 at 11:55:49 AM UTC-5, William Wright wrote:
Hi everyone,

I'm using Bowtie2 with RSEM. Illumina RNA-seq (total, not mRNA-enriched) paired-end . My transcriptome mapping rates are around 55% for all samples. I also need genome mapping, and the "--output-genome-bam" is very handy. However, I can't see the mapping rate for the genome. 

Additionally - I'm a bit worried because I used STAR on these samples in the past to map to the genome (Hg19 then) with >90% mapping. 
I'm not sure if this is acceptable or normal or if I should continue or stop the analysis. 

Furthermore - I tried Salmon and the transcriptome mapping was also around 55% - so I don't think it's RSEM specific. Here's my code :




$rsem -p 8 --bowtie2 --bowtie2-path $bowtie2 --bowtie2-sensitivity-level sensitive --paired-end --strandedness reverse --append-names --sort-bam-by-coordinate --output-genome-bam $indir/CLEAN_Sample1_R1.fastq.gz $indir/CLEAN_Sample1_R2.fastq.gz $reference $outdir/RSEM_hg38__Sample1.fastq.gz


Thanks!
Charlie

--
RSEM website: http://deweylab.biostat.wisc.edu/rsem/
---
You received this message because you are subscribed to the Google Groups "RSEM Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to rsem-...@googlegroups.com.
Reply all
Reply to author
Forward
0 new messages