The algorithm for the multiple alignment is as follows. First, sequence 2 is aligned with sequence 1 (i.e., block of sequences from 1-ALIGN_BLOCK). Next, sequence 3 is aligned with an average of the aligned sequences 1 and 2; i.e., the weight matrix is an average of the weights 1-3 and 2-3. For this averaging, the gap-residue and gap-gap weights are obtained from the residue-residue weight matrix file, not from gap penalties. If the corresponding weights are not in the file, they are set to the worst and best residue-residue score, respectively.
# Example for: MALIGN # This will read all sequences from a file, align them, and write # the alignment to a new file: READ_ALIGNMENT FILE = 'toxin.ali', ALIGN_CODES = 'all' MALIGN GAP_PENALTIES_1D= -600 -400 WRITE_ALIGNMENT FILE = 'toxin-seq.pap', ALIGNMENT_FORMAT = 'PAP'