AMBER Archive (2008)

Subject: Re: AMBER: silhouette width/coefficient after clustering by PTRAJ9.9

From: Jianyin Shao (jyshao2004_at_gmail.com)
Date: Fri Aug 01 2008 - 14:43:14 CDT

Hi, Dong and Kevin,

I am sorry that I sent you the wrong script. Please discard the Newick.pl
script. Instead use the Newick-step.pl I attached this time.
The script requires a ClusteringMerging.txt or ClusteringSplitting.txt file
and total number of frames in the trajectory, also you can provide a number
n to show the last n steps of merging. The previous Newick.pl generates all
steps, which the plottree cannot handle.

Best,

Jianyin

On Fri, Aug 1, 2008 at 3:39 AM, Kevin Wu <kjwu_at_ucsd.edu> wrote:

> Hi Jianyin,
>
> Recently Dong Xu contacted you about plotting denograms from ptraj
> clustering outputs, he forwarded it for me to figure it out. You told him to
> use the "Newick" perl script to get it into the PHYLIP format. I looked at
> "Newick.pl -h" and I tried to run the command:
>
> ./Newick.pl -in ClusterMerging.txt -out Newick.txt -mode merge -count 1572
>
> There was no error message outputted and Newick.txt didn't look out of the
> ordinary. The output Newick.txt is attached. However, when I submit to tree
> plot web service, "http://www.bioinformatics.nl/tools/plottree.html ", in
> your ptraj clustering paper I get the error: "Premature end of script
> headers: showtree". Am I doing something wrong?
>
> I searched for another plotting program called njplot to see if it was just
> the server, and the program told me that its not a "correct tree". Let me
> know what I'm doing wrong. Thanks for your time!
>
> Regards,
> Kevin
>

On Tue, Jul 29, 2008 at 1:40 AM, Dong Xu <d1xu_at_ucsd.edu> wrote:

> Hi Jianyin and Tom,
>
> I was wondering if there's tutorial about how to plot the dendrogram
> from ptraj clustering outputs. I see you used
> http://www.bioinformatics.nl/tools/plottree.html in the ptraj
> clustering paper. But I still can't figure out how to use that web
> site and obtain PHYLIP format from ptraj outputs. Could you shed some
> light on this topic?
>
> Thanks a lot in advance,
>
> -DX
>
> On Tue, Jul 8, 2008 at 9:45 PM, Jianyin Shao <jyshao2004_at_gmail.com> wrote:
> >
> >
> > On 7/8/08, Dong Xu <d1xu_at_ucsd.edu> wrote:
> >>
> >>
> >>>
> >>> The syntax for verbose is "verbose" followed by an integer. The purpose
> >>> is to let the developer check some internal information during the
> >>> clustering. Using "verbose" option may generate excessive output and
> cause
> >>> confusion, thus not advised.
> >>
> >> Thanks for the info. Would mind telling me what integer is commonly
> used?
> >> I'm curious.
> >
> >
> > You can try 6 or 7.
> >>>
> >>> Readmerge does not require the file PairwiseDistances to be present. It
> >>> will calculate the (rms-fitted) distances between frames, but will not
> try
> >>> to rewrite a new PairwiseDistance file.
> >>
> >> Do you Readmerge will not overwrite PairwiseDistance file if it's ready
> >> present in the current folder?
> >
> >
> > If a correct PairwiseDistances file exists in the current directory, the
> > Readmerge will read it. If not, Readmerge will calculate the distance on
> the
> > fly, but it will not write or overwrite a PairwiseDistances file.
> >
> >>>
> >>>
> >>> I don't know how to estimate the memory usage in Perl. The memory
> >>> management in Perl is mostly automatic, that is why I feel it is easy
> to
> >>> write a script. Maybe someone else can help me with that. I will let
> you
> >>> know when the silhouette function is fully incorporated in the ptraj.
> >>
> >> Ok, just FYI. I ran it on a new machine and it turned out using more
> than
> >> 10.5GB ram. I have another larger system that ended up using 18GB ram
> with
> >> this perl script.
> >>
> >> Thanks,
> >>
> >> Dong Xu
> >>
> >>>
>


-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
      to majordomo_at_scripps.edu