Comments on Assignments #7:

Most of the Macs in the Maclab do not have enough memory to run the problem in assignment6 3B.�

The two outfiles are here:� without and with ASRV

> When I used treeview (version 1.6.1) to open testseq5.phy or

> testseq5b.phy which were got by using Clustalx, I can't open them

> on my PC, the screen shows "syntax error 715", could you tell what

> may cause that and how to solve it

The files ending on .phy are sequence alignments NOT tree files.� The trees to use as a starting point were linked further down in the assignment.� In general, if you cannot open a file, check it out in a text editor.� In this case you used clustalx to transform the format of the alignment (from clustal format to phylip format) - you did not use it to create a tree.�

> When I do the question3 (part A), you write "calculate branch

> lengths and likelihood values for the different trees", does that mean

> to change "computer clocklike branch lengths" (z) from No to Yes.

No, you don't want to assume a molecular clock, you just want to calculate branchlengths and their confidence intervals for a tree topology you enter.�

> And I save the trees you gave to us as textfile (tree5b.txt) and delete

> ' in the textfile you mentioned, after that I run the Puzzle for

> question3, and "enter the file name for the user trees" tree5b.txt or

> test1.phy (one the tree I saved), but I can't get results from "outfile,

> outlm, outdist", I don't know why. How can I solve this problem?

On which machine are you running puzzle?� If you do it under windows, you might want to start a DOS window, change to the PUZZLE directory using the DOS command �cd�, and start the program by typing puzzle.� In windows, if you click on the puzzle icon, it runs in a DOS Window, but the window closes as soon as the program is finished.� If the program finishes with an error message, you never see the message, because the window closes immediately after displaying the message.� If you start Puzzle from within a DOS window, it stays open until you close it.�

The reason you cannot open outfile might be there is nothing in it because the program aborted with an error (due to wrong format, or out of memory...).� To solve this see above comment on windows.� If the program finishes normally (it tells you in the window), you should be able to open the oufile with any text editor (notepad, word, simple text).�

The outdist file also is a textfile (the same distance matrix is also displyed in outfile).� The outml.eps file contains the output from the ml-mapping (i.e. it is only generated if you do ml mapping as in task#2.� It has the same info as the outfile but as a much nicer graphic.� You can open it with any program that reads postscript.� I usually use Acrobat destiller (part of the acrobat package), but any number of graphics programs will do. - But you really can ignore this, as the same info is also in outfile.���

When I use Dos to run puzzle, and after enter a file name for

user trees, if enter *.txt (for tree), the window shows "unable to

proceed (sequences missing in tree)"; if enter *.phy, the window

shows "there is ' in the file". I don't know what I should do then.

When asked for the sequences the file should be in phylip format (*.phy).�

When asked for the usertree, the file should contain a text file with the tree in parenthesis notation.� e.g., in case you want to address the question if the red algae group in a clade with the green algae the tree file containing two trees: would look like this:

2

(Giardia,(Trypanosom,Trichomona,Plasmodium,Entamoeba,((Dictyostel,((Nosema,(Schizosacc,(Neurospora,(Candida,(Eremotheci,Saccharomy))))),(Ascidia,(Homo1,Bos1,A1Gallus),(Manduca,Drosophila)))),(Cyanidium,((Scherffeli,(1Acetabula,2Acetabula)),(Beta,(Vigna,Gossypium,(Daucus,Brassica,Arabidopsi))))))));

(Giardia,(Trypanosom,Trichomona,Plasmodium,Entamoeba,(Cyanidium,(Dictyostel,((Nosema,(Schizosacc,(Neurospora,(Candida,(Eremotheci,Saccharomy))))),(Ascidia,(Homo1,Bos1,A1Gallus),(Manduca,Drosophila)))),((Scherffeli,(1Acetabula,2Acetabula)),(Beta,(Vigna,Gossypium,(Daucus,Brassica,Arabidopsi)))))));

However, the idea was for you to arrive at the two topologies yourself using the tree editor that is part of treeview. Again, you need to delete the ' if you copy and paste the trees from treeview into a text editor.�

Of course, you need to make sure that the trees you supply, and the alignment you use, have the same names for the sequences.� If , for example, in making the alignment file you left in some prokaryotic sequences, these are not part of the tree, and the program will tell you so.�� Could it be that you use testseq5 for the alignment, and a tree that corresponds to testseq5b.phy, i.e. a data set with many fewer species?�