Greetings:
 
Thank you to everyone who took the time to reply to my recent posting asking
for programs or methods for concatenating multiple sequence alignments into
one overall alignment. Several individuals asked that I share the results of
my query regarding concatenating multiple sequence alignments (MSAs). First
and foremost, many reminded me to make sure that the sequences in each MSA
were in the same order (i.e., if MSA one contained sequences from Human,
Chimp, Mouse, and Chicken in that order, the sequence in all other MSAs had
to be in the order Human, Chimp, Mouse, Chicken). Second, for several of the
most efficient methods to work, the sequences in the different MSAs need to
have the same sequence ID as most of the programs use this information to
match and stitch together the sequences.
 
The most common and generally simplest recommendation was to append each MSA
one below the other in one file and then input this into PHYLIP or PAUP as
interleaved sequence. For PAUP (and MacClade and Mesquite and MrBayes) this
will require that the file be a nexus-format file. For PHYLIP the file will
need to be in the standard PHYLIP format. With this method the program will
concatenate the sequences in the process of reading them into the analytical
pipeline. PAUP and MacClade (and possibly Mesquite?) can then export the
processed input file as sequential data and in several formats.
 
Another sequence analysis program recommended by several responses was
DAMBE, which has a concatenation function among its features. DAMBE can be
downloaded from http://dambe.bio.uottawa.ca/software.asp. DAMBE is a Windows
program, but presumably it can work in an emulator (Virtual PC) or virtual
machine (Parallels desktop).
 
Still more individuals recommended using a Perl script and/or Perl and
BioPerl modules to perform the concatenation. If one is running Mac OS X
then perl is an available option.
 
Another interesting alternative was to use Excel. One can open the alignment
file as text so that the sequences are in individual cells. Though not
absolutely required the individual MSAs can then be copied and pasted into
one spreadsheet. Then the function concatenate() can be used to concatenate
the cells with the appropriate sequences. While itıs not necessary to have
the sequences in the MSAs in the same order with this method it would be
much easier to implement if they were.
 
Other programs mentioned that can perform concatenation were ProSeq v 3.0
(latest development version can be downloaded from
http://helios.bto.ed.ac.uk/evolgen/filatov/proseq.html), the bioinformatics
workbenches CLC Gene, Protein, and Combined (fully functional demos
available from http://www.clbio.com), Genious (7-day trial available from
http://www.genious.com), the Se-Al alignment editor ( available from
http://evolve.zoo.ox.ac.uk/software.html?id=seal), and the BioEdit Windows
biological sequence editor (available from
http://www.mbio.ncsu.edu/BioEdit/bioedit.html).

Cheers,
Kurt Wollenberg, Ph.D.
Phylogenetics and Sequence Analysis Consultant
Biocomputing Research Consulting Section
Bioinformatics and Scientific IT Program (BSIP)
NIH/NIAID/OTIS
Contractor, MSD Inc.

wollenbergk@niaid.nih.gov