Universit at Bielefeld Forschungsschwerpunkt Mathematisierung | Strukturbildungsprozesse on Simultaneous versus Iterative Multiple Sequence Alignment on Simultaneous versus Iterative Multiple Sequence Alignment

One of the main problems in computational biology is the construction of biologically plausible alignments for given sequence families. Many procedures have been developed for this purpose; most are based on clustering the sequences hierarchically prior to the computation of their alignment and then proceeding recursively. In this paper, we compare results obtained with a new algorithm for simultaneous multiple sequence alignment, the so-called Divide & Conquer Alignment procedure (DCA), with results obtained with iterative procedures. DCA allows one to simultaneously align relatively large families of sequences using computa-tionally eecient heuristics which { relative to the sum-of-pairs scoring function which DCA accepts as the standard of truth { give rise to good (though not necessarily optimal) alignments. As expected, the simultaneous alignments often prove superior in picking up biologically important signals contained in a family of highly diverged sequences, which are sometimes overlooked by successive pairwise alignment.