Mapping to single sequence

Question

[PriyaD]

Hello All,

As a part of one experiment we generated Few hundred thousands reads from a synthetic RNA ~70nt in size (Known sequence ordered from IDT). But prior to the cDNA synthesis this RNA molecule was fragmented in approximately 10-40nts (As a part of protocol). Now we want to map them back to the known sequence we have (Just one sequence). In this regard I need your help, Which approach would be best to map these reads back to the known sequence. Can we create a custom database that can be used to mappers like BWA or Bowtie2 or any other mapper that you think would be a good fit for this approach. If we can create a custom database with single sequence which tool can help me in better way.

Any help/suggestions are welcome.

Thank you

Answer 1

[GenoMax]

You will want to be careful about removing adapters from the sequence (I assume libraries were made for Illumina sequencing?) before you align the data. Since you expect the inserts to be short you will want to use an alignment program that does ungapped alignments (e.g. bowtie v.1.x) or use bbmap.sh with ambig=all vslow perfectmode maxsites=1000.

Creating an index with just one reference sequence should be done as normal.

[PriyaD]

@Genomax: Thank you very much for the reply, Yes the libraries were made for Illumina sequencing, I did remove the adapters and Re-Qced them with FastQC.

I did want to ask a follow up question regarding the QC. Is it possible that we might have leftover adapters presence even the FastQC is not showing them in the report and how much they might affect the mapping in this scenario or is it depends on the mapper I will select to map.

[GenoMax]

Even if you have some sequence left over it should be soft-clipped by the aligner.

[PriyaD]

Thank you very much for your help, I will try and update on this thread