First, I would like to thank you for making MAFFT available to the
research community. 

I have several comments regarding MAFFT:

1) It should be a high priority to update the manual page to include all
options, clarifying the meaning of the aliases. 
This would make a huge difference in knowing how to implement
MAFFT as part of a larger bioinformatics system.

2) It would be useful if the actual commands used in the online version
were documented somewhere. 

3) Why are the symbolic links such as einsi, linsi etc. not included in
mafftdir/bin directory, in the binary distributions? It is unusual to have
a single script (ie. mafft)
called by different aliases, to give different functions. I was only able
to figure this out after looking at the mafft script, and also  looking
at the Makefile in the source version. 

4) There should be a separate output file with a report on the particulars 
of a given alignment. These might include the parameters, the algorithms,
information about the alignment such as length, number of sequences. In particular,
it would be very useful to have the alignment score for an alignment, so that
alignments can be compared.

5) In the Online version, theree are 3 choices for the scoring matrix for
nucleotide sequences: 200PAM/k=2, 20PAM/k=2, 1PAM/k=2. How can these be
specified at the command line?

I am planning on including MAFFT in the next release of my BIRCH bioinformatics system.

home.cc.umanitoba.ca/~psgendb
 
MAFFT would be run from our BioLegato graphical interface:

http://home.cc.umanitoba.ca/~psgendb/tutorials/bioLegato/overview/overview.html

A Development version of BIRCH including MAFFT
can be downloaded for Linux or MacOSX using the
GetBirch installer, Advanced install, at

http://home.cc.umanitoba.ca/~psgendb/birchhomedir/FTP/BIRCH/getbirch_site/main.html