“Tree analysis using New Technology”
Goloboff, Farris, & Nixon, 2003
Version 1.5 now available
with a full implementation of landmark data
TNT stands for "Tree analysis using New Technology". It is a program for phylogenetic analysis under parsimony (with very fast tree-searching algorithms; Nixon, 1999, Cladistics 15:407-406; Goloboff, 1999, Cladistics 15:407-428), as well as extensive tree handling and diagnosis capabilities. It is a joint project by Pablo Goloboff, James Farris, and Kevin Nixon.
The Willi Hennig Society
As of November 2007, the project was subsidized by the Willi Hennig Society, and thus the program is now made freely available, upon agreement on the terms of the license presented by TNT when it is run for the first time in a machine (a copy of the license can be found here). The license provided is a single-user license, and the most important requirements are that the program is not posted in other websites or mirrors, and that when publishing results you acknowledge having used the Willi Hennig Society edition of TNT and cite the paper describing this version of the program (Goloboff & Catalano, 2016, Cladistics, DOI 10.1111/cla.12160)
Notes to run TNT in parallel, under Linux or Mac.- You have to install PVM in your machine, and copy the binary file to the PVM directory for binaries (either to $PVM_ROOT/bin/LINUX, or LINUX64, or DARWIN). Mixing 32 and 64 bit versions poses no problem, but binaries with limits on number of taxa do not connect properly to binaries without such limit, and viceversa. To avoid mixing, you can rename the binaries of any type (and make sure you copy identically renamed binaries to the corresponding PVM directory). You should no longer have problems compiling PVM, if you download a recent version (previous versions had problems with some compilers, but they have been fixed now --thanks, PVM guys!!).
Note on PVM and Mac 64.- The makefile provided with the PVM package does not explicitly specify the -m64 switch (which, in at least some compilers/installations, is not the default, even in 64 bit machines). The 32-bit daemon communicates correctly with TNT, but if you want to use PVM to transfer a LOT of data, having a 64-bit daemon may be good to let it access more memory. If you want to compile PVM for 64-bits, copy this makefile to your $PVM_ROOT directory, then type "make clean" followed by "make install".
Updates.- Updates are done regularly (often every 2 months or so). The last update was done on August 18, 2016. A list of recent additions and bugfixes is maintained here.
A simple, one-shot analysis is now included in the aquickie.run script, which comes with all TNT versions. This runs a simple analysis with a search for most parsimonious trees (or asks whether you want to do a consensus estimation, if the data set is large), then calculates group supports using resampling. To run this under Windows, you first open your data set, and then open aquickie.run. Alternatively, to run the data contained in file "filename.tnt" open a DOS shell, and type:
tnt filename.tnt ; aquickie ; [enter]
Linux/Mac.- You have to open the data file yourself, and then run the script. To do this, enter TNT and then type at the command line:
proc filename.tnt ; aquickie ; [enter]
The script will provide you with a few choices, then run the analysis automatically.
The program GB2TNT (Goloboff & Catalano, 2012, Cladistics 28: 503–513) provides an easy way to create data sets for TNT from GenBank data (including all the taxonomic information from the GenBank files, so that it is easy to diagnose results and color different taxonomic groups). The program and documentation (still in beta; current version is 0.75) can be downloaded here.
The main documentation is in the file tnt.htm, and recent additions or bugfixes are in the file BugFixes.htm; both of these are contained within all the packages (except the character-mode Windows version). On-line help on all TNT commands can be obtained with the help command. A general description of TNT version 1.0 for Windows is in a QuickTutorial (a PowerPoint presentation; this is probably the easiest way to get started). Some example files for scripting come in the self-extracting files zipdruns.zip or tnt_scripts.zip, and additional scripts are here (with full documentation of the scripting language in a pdf file). Instructions to view large trees in Windows and process the taxonomy contained within taxon names (as in Goloboff et al., 2009, Cladistics 25:211-230) can be found here
Keep in mind that freezes/crashes might occassionally occur, and save results frequently. The program is intended only for strictly academic use, and is provided "as is," with no express or implied warranties. None of the authors of TNT is responsible in any way for any problems the program causes to your computer, your data, your career, or your life. A copy of the license agreement can be found here
Bug reports are appreciated, but please keep in mind that a bug report, to be of any use, must include the version of TNT (including date), the complete data file, and all the commands necessary to show the bug (or, at least, the exact sequence of menu operations that consistently leads to a crash, freeze, or miscalculation). Send bug reports to Pablo, with "tnt bug report" in the subject.
To properly view the trees drawn by TNT with extended ASCII characters you will need to install the Tred.ttf font (available here). In Windows, you can just copy it to the Fonts directory (or install from control panel). In Linux or Mac, you will have to copy Tred.ttf to the directory where other True Type fonts (*.ttf) are located (probably something like /usr/X11R6/lib/X11/fonts/TTF/ in Mac, or /usr/share/fonts/truetype/ in Linux). The second step you need to do, for each file to view, is to convert it to unicode characters, using the converter to unicode (available for Mac Intel, Linux 32, and Linux 64):
tounicode tnt_output_file unicode_file [enter]
Then, you can open unicode_file with OpenOffice, or the like, select all text, and change font to Tred. The trees will then look ok.