Next: Description of the Up: Introduction Previous: Predicting the likely

Multiple alignment strategy

  1. All pairwise comparisons between the sequences are performed. The information obtained may be used to construct a dendrogram that visualises the groups of similar residues.

  2. Given a group of similar sequences, the order in which they should be aligned is determined (ie most similar pair through to least similar).

  3. The multiple alignment algorithm is then applied to the sequences. Firstly the most similar pair are aligned, then the next most similar sequence is aligned to the alignment of the most similar pair. Then the next most similar sequence is aligned and so on..... An optional process is to re-align each sequence with the completed alignment in order to refine the alignment.