Progressive Alignment

Progressive Alignment is a heuristic strategy for multiple sequence alignment that builds an alignment incrementally, starting with the most similar pair of sequences and adding others one by one. The order is determined by a guide tree constructed from pairwise distances. This approach, used by Clustal and early versions of MAFFT, is fast but greedy: errors in early alignments are locked in and propagated to later sequences. Progressive alignment assumes that evolutionary similarity is a reliable proxy for alignment order, an assumption that fails for highly divergent sequences or those with domain rearrangements.

Progressive alignment is the algorithmic equivalent of building a bridge by starting from both shores and hoping the middle meets. It usually does, but when it doesn't, the gap is catastrophic.