Aligning miRNA reads to miRbase

Question

[glady]

Hello.,

Recently I have got some miRNA reads from the small RNA seq data. I have filtered them for quality and removed the adapters with FASTX. I want the align the reads to the miRbase by using SHRiMP.

1. Which SHRiMP parameters should I consider while aligning the reads to miRBase?
2. Which miRbase file mature/precursor file should I use as reference?
3. In there a need to create an index for the reference miRBase file before aligning the miRNA reads to it?

Thank you.

[Buffo]

I recommend you to use both, mature and hairpin because you may found new miRNA variants or "isoforms". And also, recommend you to use bowtie (no bowtie2 or blast) to align them, it is easy to handle and filter multihits.

[glady]

Thank you for your reply. But how can I use both the mature & hairpin files as a reference. Since, for preparing the index, I can only pass on one file.

For example, I prepared the reference file in SHRiMP by using this cmd: media/SHRiMP_2_2_3/utils/project-db.py --seed 00111111001111111100,\ 00111111110011111100,00111111111100111100,00111111111111001100,\ 00111111111111110000 --h-flag --shrimp-mode ls /project/SHRiMP_ref/miRBase_22/mature_ref_hsa.fa

[Buffo]

$cat hairpin.fa mature.fa > both.fa

[glady]

Thank you for the suggestions. How can we get the read count from the final SAM output file?

Answer 1

[karthic]

I guess performing a blast with task as short blast and setting as ungapped will do. It depends upon what you are looking for, it is common to go for the mature miRNAs from the mirbase. You can even download mirnas for the selected species of your interest. If you are going to do blast, index file is not required to create, but if you are going to align using bowtie or bwa, then you need to create an index file.

Thanks,

KK

[glady]

Which file should we consider to create an index? miRbase mature or hairpin file? In the above comment, I used an miRbase mature file of homo sapiens.

[karthic]

I have used mature miRNAs.

[glady]

What are the parameters you used for mapping? Because I'm getting very low mapping percentage - 2.77%. the command which I used is;

/SHRiMP_2_2_3/bin/gmapper-ls -L /SHRiMP_ref/mature_ref_hsa-ls --ignore-qvs -Q /sample.fastq --mode mirna >sample.sam 2>sample.log