Fig. 1

Gapped sequence alignment methods. Each plot shows a schematic dynamic programming table for a local alignment method. The gray area depicts the explored portion table, i.e. the part that is actually computed. The ‘+’ marks a k-mer match position. The red line indicates the best alignment. A Rigorous sequence alignment. The complete table is explored. Hence no match position is required as seed. B Alignment with X-drop criterion. Exploration of the table is terminated at positions, where the alignment score drops X below the score of the best alignment seen so far. In consequence, the shape of the explored area is dependent on the sequences. C Banded alignment. Table exploration is restricted to a diagonal band, i.e. only a certain number of gaps is allowed in either sequence. In Biotite the dynamic programming table is indented to reduce memory requirements by removing a large part of the unexplored area