Hi there,
I'm using miRDeep2 in order to detect and quantify miRNA in smallRNA-seq data. I'm getting some trouble with the quantification module. I use mice data and the hairpin.fa and mature.fa files from miRbase (v21). When I run the quantifier module with this command line:
quantifier.pl -p hairpin_ATCGN.fa -m mature_ATCGN.fa -r reads.fa -t mmu
I got this error
Error: No entrys for species "mmu or mmu" found in file hairpin_ATGCN.fa
Please make sure that the given species argument matches the species id in your file hairpin.fa or say none
I already removed line containing non [ATCGNactgn] characters. When I do
grep "mmu" hairpin_ATGCN.fa
I got many lines like this:
`>mmu-mir-8117-MI0026049-Mus-musculus-miR-8117-stem-loop
mmu-mir-9769-MI0031121-Mus-musculus-miR-9769-stem-loop mmu-mir-8108-MI0026039-Mus-musculus-miR-8108-stem-loop`
I don't understand where is the problem with my files.
I also checked my mature_ATGCN.fa file and got:
`>mmu-miR-8096-MIMAT0031398-Mus-musculus-miR-8096
mmu-miR-8097-MIMAT0031399-Mus-musculus-miR-8097 mmu-miR-8098-MIMAT0031400-Mus-musculus-miR-8098`
Does anyone already solve this problem?
Thank you for your help
Sebastien
*Alternative to mirdeep;
You can also "quantify" miRNAs expression with Bowtie1 (No mismathches allowed), mirdeep uses the same method. You have to make the bowtie index (database) and then, map you reads to the reference (index). Sort the sam file from the mapping and then using samtools idxstats you can get the "quantification".
In the miRDeep2 paper, the authors say they intersect mapping file coming from the alignment of the reads against a precursor file and the mature miRNA against the same precursor file. This step is used in order to detect the precursor from which the reads come. Do you think sorting unique lines could be enough to do that?
No, reads from miRNAseq have errors (in sequence and lenght) so you need to use the entire files for better results, why you don´t want to use them? It doesn´t take long computing time.
I didn't say that. I was talking about the identification of miRNA's precursors. Which file are you talking about? I don't follow your idea
Ok, identify miRNA precursor is a very complicated issue because the presence of isomirnas, and I don´t know if using mirdeep you will get that, I don´t think so. If I where you, probably I will start identifying the most interesting expressed mature miRNAS on my samples and then try to identify the probable precursors (harpin sequence) by mapping to the reference genome and testing the coverage with longest reads (from your own mirna-seq) for this regions.
Do you know another tool that enable the miRNA identification and the isomirna identification? I heard about miRspring to do that. Do you know that tool?