fixed states optimization

1 view
Skip to first unread message

Meifeng

unread,
Jun 2, 2008, 6:30:54 AM6/2/08
to poy4
Dear all,

I couldn't manage to transform more than one read file into fixed
states.
It only works when I put for instance 18S+morphology.
I tried:
- read one sequence file, transform to fixed states, then read another
file and do transforming--> doesn't work, it gives a
"not_found" (although the first input file has been transformed)
- read all sequences files first, then perform them to fixed states-->
"not_found".

Is the fixed_states applicable only for one file in one analysis?

Thank you
isabelle

Andres Varon

unread,
Jun 2, 2008, 4:02:27 PM6/2/08
to po...@googlegroups.com
Hello Isabelle,


On Jun 2, 2008, at 6:30 AM, Meifeng wrote:

>
> Dear all,
>
> I couldn't manage to transform more than one read file into fixed
> states.
> It only works when I put for instance 18S+morphology.
> I tried:
> - read one sequence file, transform to fixed states, then read another
> file and do transforming--> doesn't work, it gives a
> "not_found" (although the first input file has been transformed)
> - read all sequences files first, then perform them to fixed states-->
> "not_found".

I can't reproduce this, could you please describe the data and post
the script that present the error? There is no need to access the
actual data, but I will need to know how many fragments, and the
precise script that you used to read the data in.

>
>
> Is the fixed_states applicable only for one file in one analysis?

No, in principle any number of characters could be transformed into
fixed states. What you are experiencing is a bug. The portions
involved in this functionality where probably the last bit of non-
trivial functionality that was changed in POY before the release, so
it is a good candidate to find some trouble.

best,

Andres
>
>
> Thank you
> isabelle
> >

Meifeng

unread,
Jun 2, 2008, 6:09:33 PM6/2/08
to poy4
Hello Andres,
Thank you for you reply. here is what I've done:

- first try:

wipe ()
set (log: "FSOcomb.log")
(*5 files with molecular data only*)
read ("/Users/Zouzou/Desktop/SequencesEriococcidae/18S#.txt","/Users/
Zouzou/Desktop/SequencesEriococcidae/28S.txt", "/Users/Zouzou/Desktop/
SequencesEriococcidae/coisequences.txt", "/Users/Zouzou/Desktop/
SequencesEriococcidae/COII.txt","/Users/Zouzou/Desktop/
SequencesEriococcidae/EF.txt")
select (terminals, files: ("/Users/Zouzou/Desktop/
SequencesEriococcidae/18Sfamille.txt"))
set(root: "Planococcus_ficus")
transform(fixed_states)

-----> error message in POY: "not_found"


- second try:

wipe ()
set (log: "FSOcomb.log")
(*molecular data dile*)
read ("/Users/Zouzou/Desktop/SequencesEriococcidae/18S#.txt")
select (terminals, files: ("/Users/Zouzou/Desktop/
SequencesEriococcidae/18Sfamille.txt"))
set(root: "Planococcus_ficus")
transform(fixed_states)
------> works

read ("/Users/Zouzou/Desktop/SequencesEriococcidae/28S.txt")
transform (fixed_states)
-------> error ("not_found")

Isabelle


Andres Varon

unread,
Jun 2, 2008, 7:11:24 PM6/2/08
to po...@googlegroups.com

On Jun 2, 2008, at 6:09 PM, Meifeng wrote:

>
> Hello Andres,
> Thank you for you reply. here is what I've done:
>
> - first try:
>
> wipe ()
> set (log: "FSOcomb.log")
> (*5 files with molecular data only*)
> read ("/Users/Zouzou/Desktop/SequencesEriococcidae/18S#.txt","/Users/
> Zouzou/Desktop/SequencesEriococcidae/28S.txt", "/Users/Zouzou/Desktop/
> SequencesEriococcidae/coisequences.txt", "/Users/Zouzou/Desktop/
> SequencesEriococcidae/COII.txt","/Users/Zouzou/Desktop/
> SequencesEriococcidae/EF.txt")
> select (terminals, files: ("/Users/Zouzou/Desktop/
> SequencesEriococcidae/18Sfamille.txt"))
> set(root: "Planococcus_ficus")
> transform(fixed_states)

I can't reproduce it yet, but here is one question: what happens if
you do the following (which is equivalent to your script):

wipe ()
set (log: "FSOcomb.log")

read ("/Users/Zouzou/Desktop/SequencesEriococcidae/18S#.txt","/Users/
Zouzou/Desktop/SequencesEriococcidae/28S.txt", "/Users/Zouzou/Desktop/
SequencesEriococcidae/coisequences.txt", "/Users/Zouzou/Desktop/
SequencesEriococcidae/COII.txt","/Users/Zouzou/Desktop/
SequencesEriococcidae/EF.txt")

transform(fixed_states)


select (terminals, files: ("/Users/Zouzou/Desktop/
SequencesEriococcidae/18Sfamille.txt"))
set(root: "Planococcus_ficus")


Does it work then? If yes, could you verify that the input data is
correct in report (cross_references)?

best,

Andres

Meifeng

unread,
Jun 2, 2008, 7:29:49 PM6/2/08
to poy4
Andres,

I've tried:
read()
transform(fixed_states)

the log gives:
transform (fixed_states)
Starting Transforming
Transforming 0 of 1 of transformations applied
Error:
Not_found


and
read()
set(root:"")
transform (fixed_states)

same thing for this one.

So the cross reference doesn't have any influence on it.

I've noticed in the log that input files are diagnosed, but is it
normal that I have for each file:

The file /Users/Zouzou/Desktop/SequencesEriococcidae/
EF.txt contains sequences of 12 taxa, each sequence holding
1 fragment.

Starting Diagnosis
Diagnosis 0 of 0 Recalculating trees
Finished Diagnosis
Diagnosis 0 of 0 Finished

Do you think it can be linked? Although I am sure it does the same
thing when I only input one file... and the transform step works.

isabelle




Meifeng

unread,
Jun 4, 2008, 12:14:44 PM6/4/08
to poy4
Hello Andres,

I talked to Cyrille, and we found out about my problem. I was using an
older version of POY4 and the latest one released must have fixed the
problem because I can now run the tranform (fixed_states).

Thank you for your help anyway!
Cheers,
Isabelle
Reply all
Reply to author
Forward
0 new messages