next up previous
Next: Tree of Minimum Gap Up: Semi-automatic multiple alignment Previous: Semi-automatic multiple alignment

Multiple alignment with considerably diverged sequence data

Here semi-automatic multiple alignment method is proposed. The basic idea is as follows:

1.
Sequence data set is classified under certain criteria such as sequence lengths and composition.

2.
Multiple alignment is performed locally in each classified data set.

3.
The multiplly aligned data sets are joined sequentially. This operation is performed by profile option of ClustalW [Thompson et al., 1994].

4.
A guide tree is constructed.

5.
Multiple alignment of whole data set is carried out following the guide tree.

6.
Steps 4 and 5 are iterated until no improvement is observed.

Programs make_gap_matrix and make_freq_matrix have been provided for the step 1.



Satoshi OOta
1999-03-06