miRWoods is a software designed for the sensitive detection microRNAs, including those with only one read. miRWoods uses two random forests. The first random forest referred to as the miR Product Random Forsest (MPRF) assembles read stacks into products and scores them. Products which don't pass through the MRPF are filtered out, and those which do are combined with surrounding products and folded to produce hairpins. Each hairpin is scored and then passed through a Hairpin Precursor Random Forest (HPRF) to obtain the final results.
miRWoods Requires the Vienna RNA fold package and Perl Library which may be downloaded here:
https://www.tbi.univie.ac.at/RNA/
Cutadapt is used for adapter trimming and may be downloaded using the following command:
pip3 install --user --upgrade cutadapt
Bowtie is used for read mapping and may be downloaded at the following address:
http://bowtie-bio.sourceforge.net/index.shtml
miRWoods uses Bio::DB::Sam which requires an earlier version of samtools. This version may be downloaded at the following address:
http://sourceforge.net/projects/samtools/files/samtools/0.1.18/samtools-0.1.18.tar.bz2
When installing samtools, edit the Makefile so that CFLAGS= '-g -Wall -O2 -fPIC #-m64 #-arch ppc'
The following perl libraries are required and can be downloaded using CPAN:
- Bio::DB::Sam
- Statistics::R
- Math::CDF
miRWoods also uses some R code which may be downloaded here:
https://www.r-project.org/
R will require the following libraries which can be downloaded using CRAN:
- ROCR
- randomForest
- mlbench
- Add the Scripts directory to your path
- Add the Perl Modules directory to your PERL5LIB path
The software takes it's inputs in the form of bam files. These may be created from fastq files using the following command:
createBam.pl < fastq file > < 3' adaptor > < min avg quality score > < bowtie index >
Here < min avg quality score > is set to the average quality score requred for a read to be kept.
We set < min avg quality score > to 30 for our experiments.
After creating each of the bam files you should place them into a tab delimited bam file list (see example_bamlist.txt for an example.)
The easiest way to pass inputs to miRWoods is through a config file (see example_config.txt) The config file includes the following information:
- scriptDir - the location of the miRWoods/Scripts directory
- hairpinRF - the location of the HPRF model file
- productRF - the location of the MPRF model file
- SizesFile - A tab chrom.sizes file containing each chrom and it's size
- genomeDir - the directory of the genome (each chrom must be split into individual files)
- geneModels - a gff file containing the genome annotations
- mirbaseGff - the gff with the mirbase annotations
- bamListFile - a tab delimited file containing the list of bam files
- minReadLength - the minimum read length for a mapped read to be processed
- maxReadLength - the maximum read length for a mapped read to be processed
- HPDVCutoff - the HPDVCutoff threshold
- ARPMCutoff - the ARPMCutoff threshold
After the config file is setup, miRWoods may be run with the following command:
miRWoods -L <config file>
The following two files are output by miRWoods:
- predictedMiRs.gff - a gff file containing the hairpins and products predicted by miRWoods
- predictedMiRs_productInfo.txt - a tab delimited file containing read count information for each product.