miRNA analysis by miRDeep2

Question

[Xin]

Hey guys,

I would like to analyze a set of miRNA data by suing miRDeep2. I have three samples and each has two replicates.

I managed to find out novel and known miRNAs by using config.txt in mapper.pl command line like this: mapper.pl config.txt -d -c -i -j -l 18 -m -p genome_index -s reads.fa -t reads_vs_genome.arf So I had one reads.fa and one reads_vs_genome.arf, then I used miRDeep2.pl.

Now I would like to find expression of miRNA in each sample to compare expression of each miRNA in different samples. Should I use miRDeep2.pl for each sample and its replicate separately or there are other ways. I know miRDeep2 pipeline but I don't know how I can mange expression analysis of multiple samples with this tool.

Thank you in advance.

[Buffo]

As first approach I would recommend you doing by yourself, you can download mirna database from mirdb, and then map your reads (each sample) to the database with bowtie, after that you can count reads mapped to the mirDB with samtools idxstats or htseq, normalize and found differential expression. Manually it is the same (or very close) pipeline as in mirdeep.

[Rituriya]

Hi Xin,

I am also perplexed as to how to find common set of miRNAs across samples and their expression levels. Also, there is no direct list generated, which gives the user a common set of novel miRNAs across samples, as they have randomly generated IDs. If you get an answer, please post it here for others to know too.

Thanks, Rituriya.