Section: Software

Magnolia – multiple alignment via comparative analysis

Magnolia is a new method to construct multiple alignments of nucleic acid sequences. It exploits our ideas coming from comparative analysis presented for noncoding RNAs and for coding sequences. Nucleic acids sequences are recognized to be hard to align because similarity is often reduced at the DNA level. The idea implemented is Magnolia is to take into account the putative function of the sequences and to incorporate this functional information into the alignment. It takes as input a set of unaligned nucleic acids sequences, classifies the sequences either as coding RNAs using protea or noncoding RNAs using carnac and produces a multiple sequence alignment based on the appropriate evolutionary pattern. When sequences are predicted as coding, then the multiple alignment relies on the putative amino-acid sequences with ClustalW . When sequences are predicted as noncoding, then the multiple alignment relies on the putative conserved secondary structure with gardenia . Magnolia has been evaluated on large experimental data sets and appeared in  [41] .