Question: low number mapped reads in ribosome profiling
0
gravatar for laura.fancello
4.8 years ago by
laura.fancello0 wrote:

Dear all, I performed both genome and transcriptome mapping on ribosome profiling data and I obtain low percentages of mapped reads: 14-20% for genome and 3-8% for transcriptome mapping. 

I am aware that the short read length (27-30bp) can be responsible for that but I am still surprised for the low number of mapped reads. Has anybody experienced the same? I only found two articles in literature where they report the percentage of mapped reads and they also have very low number of mapped reads for some samples (Reid et al., J Biological Chemistry 2012 and de Klerk et al., NAR 2015). Still I am not sure if this is the most common case.

 

Some details on my mapping:

The mapping was performed after eliminating rRNA and tRNA contamination (by bowtie mapping).

For genome mapping I used Tophat allowing 2 mismatches

For transcriptome mapping I used bowtie with options -l 25 -n 1 -m 1 against a filtered gencode transcriptome containing only validated or manually annotated protein coding transcripts and in case multiple isoforms existed per gene only the longest was used.

 

Thanks,

Laura

 

ADD COMMENTlink modified 4.4 years ago by Bontus80 • written 4.8 years ago by laura.fancello0

Dear Laura,

I am facing a similar issue, albeit not exactly the same, as only around 30% of the reads after trimming and rRNA filtering are mapping uniquely to the genome, whereas overall mapping rate is quite high (= many multimappers). So have you had any luck with finding an explanation for your problem, as maybe there is an underlying cause this kind of behavior? My current strategy is to include all potentially contaminating RNAs (Mt_rRNAs, tRNAs, snoRNAs, snRNAs and pseudogenes) in a filtering step prior to genome-wide mapping. One possible reason could be overamplification during library preparation, but I doubt that this is the case for us since this is a pilot study and there was a lot of material to use.

I found some other papers mentioning uniquely and multimapping reads, one mentioning explicit numbers (quite low).

and the other presenting a bioinformatics solution for counting also multimapping reads in Ribo-seq (again low numbers for unique hits).

Best,
René

ADD REPLYlink modified 2.2 years ago by _r_am31k • written 4.4 years ago by Bontus80
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 1154 users visited in the last hour