|Sourceforge Page - Contains description, documentation and download links.
MIRA works with Sanger, 454 and Illumina data. It is 12 years in the making and still being developed.
To install we first needed to download and install GNU Flex 2.5.35 and Expat 2.0.1 libraries. Both were downloaded from sourceforge.net and installed with
./configure --prefix=/campusdata/BME235/ make make install
Once the prerequisites were installed, installation was successful with with
./configure --prefix=/campusdata/BME235 --with-boost=/campusdata/BME235 --with-boost-libdir=/campusdata/BME235/lib --with-expat=/campusdata/BME235 make make install
This added the following binaries to the bin directory
-rwxr-xr-x+ 1 mcusack bme235 3.0M Apr 11 21:18 convert_project -rwxr-xr-x+ 1 mcusack bme235 11K Apr 11 21:18 fasta2frag.tcl -rwxr-xr-x+ 1 mcusack bme235 3.2K Apr 11 21:18 fastaselect.tcl -rwxr-xr-x+ 1 mcusack bme235 86K Apr 11 21:18 fastatool -rwxr-xr-x+ 1 mcusack bme235 1.6K Apr 11 21:18 fixACE4consed.tcl -rwxr-xr-x+ 1 mcusack bme235 3.1M Apr 11 21:18 mira lrwxrwxrwx 1 mcusack bme235 4 Apr 11 21:18 miraclip -> mira lrwxrwxrwx 1 mcusack bme235 4 Apr 11 21:18 miramem -> mira lrwxrwxrwx 1 mcusack bme235 4 Apr 11 21:18 miraSearchESTSNPs -> mira -rwxr-xr-x+ 1 mcusack bme235 568K Apr 11 21:18 scftool
Slides from a presentation dated 5/17 by Michael Cusack summarizing how MIRA works mira_internals.pdf.
A list of assemblies using MIRA can be found here.
Currently just two assembies have been performed with MIRA using Pyrobaculum oguniense data.
MIRA appears to be a very good assembler for denovo data. It's strengths are that it can take in many types of data tries to use all data to the fullest while keeping quality of the results high. Based on tests there appears to be occasional problems with chimeric contigs in the results, but this was fixed by specifying more stringent settings.
Disadvantages are mainly its speed and memory requirements. Assembling Pog took 13 hours, and mira estimates it would need 1.1 TB of memory to assemble the banana slug (using data available May 2010). Documentation is also spotty and often outdated. The thesis referenced on this page is the most complete reference of how MIRA works, but it is several years (and versions) old. Some later documentation mentions that parts of the processing have changed in recent versions, but do not specify what or how.