This option allows the alignment to be performed in a manner similar to that described by Feng and Doolittle (1987). Rather than using a single order, the sequences are aligned according to a defined evolutionary tree. In my implementation, the tree can be obtained by running program ORDER, this produces two files, an order_file and a tree_file. The order file specifies an order that preserves the clustering of the sequences by similarity. The tree_file gives specific instructions to the program as to which comparisons to make and in which order.
For example, given four sequences A,B,C,D. the tree_file may specify that A and B should be aligned first, then C with D and finally the AB alignment is aligned with the CD alignment using average scores at each amino acid position.
Example command file. globin_mult_tree.com
Commands Explanation number -------- ------------------ output_file=globin_mult_tree.out 1. mode=multiple 2. matrix=file=ampsdir:md.mat 3. gap_penalty=8.0 4. constant = 8 5. consplot=mz 6. print_vertical= 7. seq_file=globin.seq 8. order_file=globin_pairs_tree.ord 9. tree_file=globin_pairs.tree 10.
This is the same as globin_mult.com with the exception of two items.
order_file defines the cluster order. This is usually different to the 'simple' order defined in globin_pairs.ord.
tree_file defines the tree_file generated by program ORDER, or manually typed in.
The example output file from this run is globin_mult_tree.out.