Skip to content

Latest commit

 

History

History
92 lines (61 loc) · 3.23 KB

TUTORIAL.md

File metadata and controls

92 lines (61 loc) · 3.23 KB
Raw

miRdeep2 tutorial

To run the tutorial please go to the tutorial subfolder.

Introduction:

The user wishes to analyze deep sequencing data mapping to a ~6 kb region on C. elegans chromosome II for known and novel miRNA genes.

Preliminary files

file description
cel_cluster.fa a FASTA file with the reference genome (this file is in fact a ~6 kb region of the C. elegans chromosome II).
mature_ref_this_species.fa a FASTA file with the reference miRBase mature miRNAs for the species (C. elegans miRBase v.14 mature miRNAs)
mature_ref_other_species.fa a FASTA file with the reference miRBase mature miRNAs for related species (C. briggsae and D. melanogaster miRBase v.14 mature miRNAs)
precursors_ref_this_species.fa a FASTA file with the reference miRBase precursor miRNAs for the species (C. elegans miRBase v.14 precursor miRNAs)
reads.fa a FASTA file with the deep sequencing reads.

Analysis

Step 1:

build an index of the genome (in this case the ~6 kb region):

bowtie-build cel_cluster.fa cel_cluster

Step 2

process reads and map them to the genome.

The -c option designates that the input file is a FASTA file (for other input formats, see the README.md file). The -j options removes entries with non-canonical letters (letters other than a, c, g, t, u, n, A, C, G, T, U, or N). The -k option clips adapters. The -l option discards reads shorter than 18 nts. The -m option collapses the reads. The -p option maps the processed reads against the previously indexed genome (cel_cluster). The -s option designates the name of the output file of processed reads and the -t option designates the name of the output file of the genome mappings. Last, -v gives verbose output to the screen.

mapper.pl reads.fa -c -j -k TCGTATGCCGTCTTCTGCTTGT  -l 18 -m -p cel_cluster \
  -s reads_collapsed.fa -t reads_collapsed_vs_genome.arf -v

Step 3

fast quantitation of reads mapping to known miRBase precursors.

(This step is not required for identification of known and novel miRNAs in the deep sequencing data when using miRDeep2.pl.)

quantifier.pl -p precursors_ref_this_species.fa -m mature_ref_this_species.fa \
  -r reads_collapsed.fa -t cel -y 16_19

The miRNA_expressed.csv gives the read counts of the reference miRNAs in the data in tabular format. The results can also be browsed by opening expression_16_19.html with an internet browser.

Step 4

identification of known and novel miRNAs in the deep sequencing data:

miRDeep2.pl reads_collapsed.fa cel_cluster.fa reads_collapsed_vs_genome.arf \
  mature_ref_this_species.fa mature_ref_other_species.fa \
  precursors_ref_this_species.fa -t C.elegans 2> report.log

Step 5

browse the results.

open the results.html using an internet browser. Notice that cel-miR-37 is predicted twice, since both potential precursors excised from this locus can fold into hairpins. However, the annotated hairpin scores much higher than the non-annotated one (miRDeep2 score 6.1e+4 vs. -0.2).