Modeltest3.7 for Maximum Likelihood test

Step-by-step Modeltest
1. Download modeltest3.7
2. Copy the modeltest3.7 folder > rename it to modeltest3.7 (remove the word 'folder')
3. Make a new folder > name it (e.g. COIfinal) > put it on the desktop
4. Align sequence in ClustalX and save it as nexus file (e.g. range 30-685) > name it (e.g. COI.nxs)
5. Copy and paste both modeltest3.7 and COI.nxs into the COIfinal folder
6. Open PAUP, execute COI..nxs
7. Then, open paupblock folder, execute modelblockPAUPb10
8. This will take several minutes > until it says completed at the PAUP software
9. Next, open modeltest3.7 folder > paupblock > you will find a file name model.scores
10. Rename model.scores > e.g. COI2015.scores
11. Copy the COI2015.scores into modeltest3.7 > bin
12. Go to start menu > All programs > Accessories > Command prompt
13. Once the program starts, type cd > then drag-and-drop your file name bin into the command
14. Make sure there is space between cd and the file-drop, Delete the marks "   "
15. Press enter
16. Type modeltest3.7.win<COI2015.scores>new file name (e.g. COIoutput.modeltest)
17. Press enter, you can see C: ....>, then close the Command prompt
18. A new output file had just been created in your bin folder
19. Open this file in text or notepad
20. Copy and paste the value : (e.g. Lset  Base=equal  Nst=2  TRatio=2.0111  Rates=gamma  Shape=0.6147  Pinvar=0.7441)
21. Run PAUP as usual

Phylogenetic tree

Constructing a phylogenetic tree

1. While  I was cleaning/editing my sequence, to construct a phylogenetic tree, I found out even 1 base can change (or make a major change) to my phylogenetic tree.

2. I also noticed that an out-group makes a lot of difference in phylogenetic tree, an outgroup using one, two or three species makes a lot of changes.

3. In addition, downloading from a genebank needs a great caution since anyone can upload it, so editing is necessary with reference from your own sample (if any).

4. To increase bootstrap value, I would throw out sample that I believe is not of quality (I used a DNA that is old of at least 1 year above)

5. So, I suggest to make a lot of trial when constructing a phylogenetic tree before proceeding with discussion.