TNT Script for Maximum Parsimony Analysis (Phylogeny)
-
Place this script and your matrix tnt file
filename
under the folder that you calltnt
orexe
file is placed -
Enter
tnt
-
Enter command
guoyi filename;
-
Extended implied weighting with K value setting (default, default K=12) or implied weighting with K value setting (default K=12) or equal weighting.
-
Search trees via implicit enumeration (ntax<=25, but not for eiw) or TBR Mult (1000 times) with branch swapping (25<ntax<75, also for ntax<=25 with eiw) or Mult with branch swapping and Xmult (use random sectorial searches, produce 50 hits to best length and stop, 10 cycles of drifting, ratchet and fusing) (ntax>=75).
-
Perform Strict consensus / Majority-rule consensus (without bremer support variations) / Half strict consensus (without bremer support variations).
-
Calculate relative Bremer support, jackknifing (1,000 times), bootstrap (1,000 times) and symmetric resampling (1,000 times).
-
Map apomorphic characters on the consensus tree.
-
Calculate TL, CI, and RI.
For Windows users
tnt run guoyi.run filename datatype weight 0/K cons resample prefix;
For Linux and Mac users
tnt run guoyi.run filename datatype weight 0/K cons resample prefix,
-
datatype should be
32
,dna
,prot
,num
or any types tnt allowed- num=number, dna=DNA, prot=protein, 32=max number allowed (default)
-
weight should be
iw
,ew
,eiw
- iw=implied weight, ew=equal weight, eiw=extended implied weight (default) N.B. ew must followed 0 (the K value position for ew iw and eiw)
-
K is
12
(default) following Goloboff et al. 2017 (Cladistics 34: 407–437)- it must more than 0
-
cons should be
str
,mjr
,hlf
- mjr=majority rule, hlf=half, str=strict (default)
-
resample should be sum of what you want
- relative bremer support (rbrs)=0.1, bremer support (brs)=0.2, jackknifing (jak)=1, bootstrap (boot)=2, symmetric resampling (sym)=4 i.e. rbrs+jak+boot+sym=7.1 (default)
-
prefix can be empty or a string
- default is empty
-
Results instructions are at the end of
tnt.log
. -
*.tre
are trees with taxname,*_no.tre
are trees without taxname. -
*.tnt.tre
can be read by TNT/WinClada directly. -
*.ctf
can only be read by TNT viashortread
. -
*.svg
is the tree file, which can be edited by inkscape -
original*
is the consensus tree. -
apo.tre
is the apomorphic character mapping tree. -
resample.tre
is the consensus tree with support, which can be read by figtree after processing by tnt2figtree. -
trees*.tre
contains he MPTs, which can be read by figtree after processing by tnt2figtree. -
resample/apo.log
contains the tree tags. -
eiw.log
contains character concavities. -
homo.log
contains homoplasy report of all characters. -
report.log
contains CI RI TL. -
winclada.ss
includes data matrix and consensus tree and it can be read by winclada directly. -
winclada.tre
can be transferred to the acceptable format for WinClada by tnt2winclada and the output file can be put into Winclada with your tnt matrix filefilename
for mapping apomorphic characters and homoplasy.
Cite this script is mandatory, list me (Guoyi Zhang) in your Acknowledgements is recommended. This script follows MIT License.
To make tnt exported trees can be readable by WinClada (Nixon, 2021) and FigTree, please check tnt2winclada and tnt2figtree in this repository.
runwincladtree.run can handle wincladtree script with input file which must be $(PREFIX).winclada.ss or winclada.tree.