Features: Build the alignment by series of pairwise alignments following the branching order in a phylogenetic tree (see Distance based methods) Build an initial alignment starting with the most related sequences Gradually add in the most distant sequences Alignments sequence-sequence, sequence-alignment, alignment-alignment Drawbacks: Greedy algorithm: Local minimum problem. Dependence on the initial pairwise alignment. Solution: Stochastic or iterative methods. The choice of alignment parameters: when adding more distantly related sequences to the msa, we are dealing with different evolutionary distances so that the substitution matrix and the gap penalties must be different on every stage.