Software

MAmBA (Methylation Assessment of microRNAs after Bisulfite Analysis)

MamBA is an analysis pipeline specifically designed to attain a (h)m5C profile of microRNAs starting from a gzipped fastq file obtained from bisulfite-treated small RNAs.

The current version of MAmBA allows one to obtain a (h)m5C profile in graphic format (.png) as well as in text format for each single microRNA.

Analysis takes about 1 minute/Million reads on a typical 8-core Desktop PC.

You will also need to download a local copy of a Bowtie2 index for the genome of the organism you want to analyse (if you want to use reference data provided with MAmBA you should download bowtie2 index hg38 for Homo sapiens, and bowtie2 index mm10 for Mus musculus)

See below for further organisms and obtaining custom MAmBA references.

Please refer to the above links for installation of each software.

Obtaining MAmBA:

MAmBA can be downloaded from github: https://github.com/flcvlr/MAmBA

To use MAmBA simply download the .tar.gz archive, extract it to a suitable location on your LINUX filesystem and run mamba.sh (to process a fastq.gz file).

Running each script with the -h option and no arguments will provide a detailed description of command line options.

MAmBA reference data:

MAmBA is provided with reference files for analysis of human (hsa) and mouse (mmu) data. In case you need to analyze a different species a script (mamba_reference.sh) to build a custom reference is included, you will have to provide annotation files and bowtie2 index for your species of interest.

Testing that MAmBA is working nicely on your machine:

To check your installation you can download test file1 and test file2

Running mamba.sh on each of these two files (with human annotation, -s hsa), you should end up with two directories which contain a summary (hairpin_subbed_sig.bed) tab delimited files with all data on the different C residues tested. Methylation profile figures for some miRNAs are contained in the subfolders "images/mathylated" and "images/non_methylated" of the output directory of each sample.

Feedback

The project is currently under development and the current version will undergo several improvements with time. Any feedback (valerio.fulci at uniroma1.it) is more than welcome, I will do the best I can to address all issues.