ncRNA database for mus

Question

[tommy]

Hello,

I am currently trying to remove ncRNA from sequencing using bowtie which requires reference sequencing. As my sequencing data from mus musculus, I found the following data sources:

1. noncode.org which claimed has one NONCODEv6_mouse.fa.gz for noncoding sequencing. However, Chrome suggested it was from a non-credential website and suggested not downloading it.
2. RNA central, by selecting ncRNA and Mus musculus, 188,951 sequences were filtered out, Should I use all of them? The results also included the AI-generated summaries, should I include them?
3. Are there any databases people normally use in order to filter out the rRNA, tRNA etc. before mapping in the bioinformatics pipeline?

I appreciate your help.

[GenoMax]

What is this experiment about? You are only referring to removing things but perhaps nothing may need to be. After alignments to the genome you can only count what you need from the alignments.

[tommy]

It's about polysome profiling. I did see some debate about whether removing such contaminants is necessary, but I am unsure in which circumstance it's necessary. If it's necessary, do you have any suggested databases?

[GenoMax]

Data from RNAcentral should be comprehensive and includes noncode.

[tommy]

Hello, Max. Do you mean that removing nc is not necessary for all RNA sequencing?

[jeevii]

Hello,

For removing ncRNA from your reads, you can try mapping to mouse ncRNA sequences which is available in ensembl database [https://ftp.ensembl.org/pub/release-112/fasta/mus_musculus/]

Try to align and exclude mapped reads from the alignment and proceed your analysis.

hope it helps!

cheers,

[tommy]

That's very helpful. Thank you, Jeevii.

Answer 1

[Jack Tierney]

As this is for polysome profiling you may want to just remove rRNA and tRNA fragments rather than all ncRNAs. LincRNAs for example are known to be translated so it may be shortsighted to remove them. The SILVA database is your best bet I described how to access sequences here.

[tommy]

Thank you for your help, Jack. May I know the reason why only rRNA and tRNA were removed?

[Jack Tierney]

My point was more that if you are to remove anything then do not remove all ncRNA as you will be discarding valid data. Categories other than mRNA (eg LincRNAs) can genuinely be translated and form polysomes. Why would you discard this valid data?

However, as GenoMax mentioned this is often not required as the impacts on data analysis are minimal but that depends on the rest of your workflow. In Ribo-Seq rRNA and tRNA reads are often discarded as they can introduce noise that may confound results. If you have long reads on the polysome fraction then this wont be an issue as the length will not permit misassignment of multimapped reads. Shorter Ribosome Protected fragments are easier to misassign due to their shorter length

[tommy]

Thank you for your help. I understand it now.