Welcome, Guest. Please Login
YaBB - Yet another Bulletin Board
 
  HomeHelpSearchLogin  
 
Pages: 1 2 
Tipdate function (Read 11026 times)
promateus
YaBB Newbies
*
Offline


I love YaBB 1G - SP1!

Posts: 7
Tipdate function
Dec 6th, 2006 at 3:07pm
 
Hi, can you write about this function and what is the best way to use it? as I have data set of different lentiviruses integrase sequenses and i try to make sense from it.

I got different results which seems not to make sense Cry

I am newbie so any help here is wellcome...

Thanks
Back to top
 
 
IP Logged
 
Sergei
YaBB Administrator
*****
Offline


Datamonkeys are forever...

Posts: 1658
UCSD
Gender: male
Re: Tipdate function
Reply #1 - Dec 6th, 2006 at 3:57pm
 
Dear promateus,

Are you talking about the standard analysis 'DatedTipsMolecularClock.bf'?
It is simply a more general implementation of Andrew Rambaut's TipDate - I would recommend you read Andrew's paper (cited by the analysis) to become familiar with the models and what can be estimated.

In what way are the results you are getting confusing? I'll try to give a specific answer if I can understand what the issue is.

Cheers,
Sergei
Back to top
 

Associate Professor
Division of Infectious Diseases
Division of Biomedical Informatics
School of Medicine
University of California San Diego
WWW WWW  
IP Logged
 
promateus
YaBB Newbies
*
Offline


I love YaBB 1G - SP1!

Posts: 7
Re: Tipdate function
Reply #2 - Dec 6th, 2006 at 4:59pm
 
Hi Sergei, So here what i did I hope that you can correct my mistakes

in the first step I selected the best model for tree using tree model and build max likelihood tree (GTR+G+I). Once the tree
was done I used the swap alogrithem of hyphy to find better tree. Then I took this tree and tried to see if the null hypothesis of strict Molecular clock validity which in my case was regected as expected when i used a single rate model (F84) for the clock tree but was not regected when I used the GTR+G+I model. On the other hand when I used the tipdate function and I put for the clock the same model that I used for the tree, the tree was regected (P=0) again.

Now my question is how I can estimate the true evolution rate/site/year when the tree is regected?

If you can highlight me on the correct procedure for this function it will be great.

Thanks
Great software  Smiley
Back to top
 
 
IP Logged
 
Sergei
YaBB Administrator
*****
Offline


Datamonkeys are forever...

Posts: 1658
UCSD
Gender: male
Re: Tipdate function
Reply #3 - Dec 6th, 2006 at 5:09pm
 
Dear Promateus,

Is your data coding? You may try to use a codon model and impose the clock (using the Local option for when prompted) on synonymous rates only (or, as an approximation, use only 3rd codon position for your nucleotide model).
Also, make sure the tree is rooted appropriately (e.g. on one of the oldest sequences).

A general procedure for dating when a simple (dated) molecular clock is violated is implemented in BEAST - a Bayesian package written by Alexei Drummond and Andrew Rambaut.

We have a simpler alternative procedure for estimating rates when a simple clock fails as well, but it is not included in the standard HyPhy distribution. I'll put together a batch file and post it here in a few days.

Cheers,
Sergei
Back to top
 

Associate Professor
Division of Infectious Diseases
Division of Biomedical Informatics
School of Medicine
University of California San Diego
WWW WWW  
IP Logged
 
promateus
YaBB Newbies
*
Offline


I love YaBB 1G - SP1!

Posts: 7
Re: Tipdate function
Reply #4 - Dec 6th, 2006 at 5:59pm
 
Thanks for the fast response,can you elaborate on how to add the synonomus rates with the tipdate.bf ?
Back to top
 
 
IP Logged
 
Sergei
YaBB Administrator
*****
Offline


Datamonkeys are forever...

Posts: 1658
UCSD
Gender: male
Re: Tipdate function
Reply #5 - Dec 6th, 2006 at 6:35pm
 
Dear promateus,

You should run the standard analysis (using the Analysis>Standard Analyses menu from the console), under the rubrik Molecular Clock, which is named DatedTipsMolecularClock.bf.

Do not use older files, because the one I cite is the most recent and maintained version.

When prompted for data type, select 'Codon', then when prompted, select one of the codon models (e.g. MG94CUSTOM), select 'Local' next (this is important to allow synonymous and non-synonymous rates to vary independently), then provide a nucleotide bias model (you can use 010010 to simply correct for transition/transversion biases). Finally, choose 'synRate' as the parameter to constrain. This should apply a dated-tips molecular clock on synonymous rates only.

Cheers,
Sergei
Back to top
 

Associate Professor
Division of Infectious Diseases
Division of Biomedical Informatics
School of Medicine
University of California San Diego
WWW WWW  
IP Logged
 
promateus
YaBB Newbies
*
Offline


I love YaBB 1G - SP1!

Posts: 7
Re: Tipdate function
Reply #6 - Dec 7th, 2006 at 9:13am
 
Thanks Sergei on your fast response you are great Wink
I want to ask if I may about the new Bf file that can deal with situations in which the clock model fail, I think it will will be graet addition to this program that from what I see is far better than many others... Cheesy

So when its going to be added?
Back to top
 
 
IP Logged
 
Sergei
YaBB Administrator
*****
Offline


Datamonkeys are forever...

Posts: 1658
UCSD
Gender: male
Re: Tipdate function
Reply #7 - Dec 7th, 2006 at 9:17am
 
Dear promateus,

Probably about 2 weeks from now.

Cheers,
Sergei
Back to top
 

Associate Professor
Division of Infectious Diseases
Division of Biomedical Informatics
School of Medicine
University of California San Diego
WWW WWW  
IP Logged
 
promateus
YaBB Newbies
*
Offline


I love YaBB 1G - SP1!

Posts: 7
Re: Tipdate function
Reply #8 - Dec 19th, 2006 at 9:51am
 
Hi Sergei, any new news on the procedure (Bf file) for estimating rates when a simple clock fails?

Best wishes
Promi

Back to top
 
 
IP Logged
 
Sergei
YaBB Administrator
*****
Offline


Datamonkeys are forever...

Posts: 1658
UCSD
Gender: male
Re: Tipdate function
Reply #9 - Dec 19th, 2006 at 9:55am
 
Dear Promi,

I will work on the analysis this week and can hopefully finish it before the holidays.

Cheers,
Sergei
Back to top
 

Associate Professor
Division of Infectious Diseases
Division of Biomedical Informatics
School of Medicine
University of California San Diego
WWW WWW  
IP Logged
 
promateus
YaBB Newbies
*
Offline


I love YaBB 1G - SP1!

Posts: 7
Re: Tipdate function
Reply #10 - Jan 10th, 2007 at 9:06am
 
Hi Sergei, sorry to bother you but is the new dating bf file avilable?

Have nice day
Promi
Back to top
 
 
IP Logged
 
Sergei
YaBB Administrator
*****
Offline


Datamonkeys are forever...

Posts: 1658
UCSD
Gender: male
Re: Tipdate function
Reply #11 - Jan 10th, 2007 at 10:29pm
 
Dear Promi,

Sorry for the delay, but I haven't had a chance to prep the file for public consumption yet (a very busy month so far!)

Thanks for you patience, it's on my list of things to do.

Best,
Sergei
Back to top
 

Associate Professor
Division of Infectious Diseases
Division of Biomedical Informatics
School of Medicine
University of California San Diego
WWW WWW  
IP Logged
 
jackson
YaBB Newbies
*
Offline


I love YaBB 1G - SP1!

Posts: 30
São Paulo, Brazil
Gender: male
Re: Tipdate function
Reply #12 - Mar 26th, 2007 at 9:30pm
 
Hi, 
I tried to use this .bf, but I've one message of error after that I chose the units of my system.
The dataset and the tree have the year date of collected, like the example on the Tipdate.

Best,

Jackson


---- RUNNING MOLECULAR CLOCK ANALYSIS WITH DATED TIPS----
The underlying methodology developed by Andrew Rambaut is discussed
in Bioinformatics 16(4): 395-399
The following data was read:
201 species:{1Sim1958,2Hum2000,3hum2000,4Hum1999,5Hum2003,6Hum2001,7Hum2000,8Hum1997
,9Bov2005,11Hum2004,12Hum2004,13Hum1995,14Hum2002,15Hum2000,16Hum2001,17Hum1997,
19Hum2004,20Hum2004,21Hum1997,24Hum1997,42Hum1997,43Hum1997,50Sim1990,51Por2004,
53Por2004,55Fel1993,56Fel1993,57Fel1993,58Fel1993,61Fel1989,62Hum1982,8247228Hum
,63Hum1997,64Hum1996,65Hum1996,67Hum1995,68Hum1995,69Hum1996,70Hum1992,71Hum1992
,72Hum1992,73Hum1991,74Hum1991,75Hum1990,77Hum1993,78Hum1990,79Hum1990,82Hum1996
,84Hum1996,85Hum1997,87Hum1996,89Hum1997,91Hum1996,95Hum1996,98Hum1996,100Hum199
6,101Hum1996,105hum1996,106Por1980,107Hum2002,108Hum1998,109Hum1997,110Hum1998,1
11Hum1993,112Hum1999,113Hum2000,114Hum1990,115Hum1998,116Hum2000,117Bov2001,118L
am1981,119Rab2001,120Rab2001,121Rab2001,122Rab1996,123Cap1998,124Hum1995,125Hum1
995,126Hum1992,127Hum1990,128Hum1991,129Hum1991,130Hum1980,131Hum1981,132Hum1983
,133Hum1982,134Hum1990,135Hum1981,136Hum1977,137Hum1980,138Hum1989,139Hum1990,14
0Hum1992,141Hum1986,142Hum1992,144Lap1978,145Hum1993,151Hum1996,152Hum1997,153Hu
m1997,154Hum1997,155Hum1997,156Hum1997,159Hum1997,160Hum1997,161Hum1997,162Hum19
97,164Hum1998,166Hum1998,167Hum1998,168Hum1996,169Hum1998,170Hum1998,171Hum1998,
172Bov1981,173Bov1973,174Bov1973,175Hum1986,176Hum1980,177Hum1984,178Hum1990,179
Hum1990,180Hum1990,181Hum1990,182Hum1988,183Hum1988,184Hum1988,185Hum1988,186Hum
1976,187Hum1994,189Hum1993,190Hum1992,191Hum1983,192Hum1982,193Hum1981,194Hum198
5,195Hum1985,196Hum1985,201Hum1989,202Hum1995,203Hum1993,204Hum1992,205Hum1981,2
06Hum1978,207Hum1994,208Hum1994,209Hum1993,210Hum1994,211Hum1994,212Hum1995,220H
um1995,231Hum1998,232Hum1996,233Hum1974,234Hum1996,235Hum1994,236Hum1997,237Hum1
997,238Por1987,239Can1980,240Equ1981,241Equ1981,242Equ1976,243Equ1975,244Por1987
,245Por1987,246Por1987,247Lap1988,248Lap1988,249Lap1978,250Lap1983,251Rot1984,25
2Por1976,253Hum1981,254Hum1975,255Hum1982,256Hum1993,257Hum1993,258Hum1981,260Hu
m1990,261Hum1988,262Hum1987,263Hum1986,264Hum2004,267Hum2004,268Hum2004,271Hum20
03,272Hum2003,274Hum2003,275Hum2003,276Hum2003,277Hum2003,278Hum2002,279Hum2002,
282Hum2002,283Hum2002,286Hum2002,287Hum2001,289Hum2001,290Lap2004,291Mur1982};
Total Sites:761;
Distinct Sites:469
What units are the dates measured in (e.g. months)? This is only used for reporting the results.years

ERROR: Input tree must contain date information in the chosen format for all sequences.
Branch lengths for internal nodes are ignored.

Back to top
 
 
IP Logged
 
Sergei
YaBB Administrator
*****
Offline


Datamonkeys are forever...

Posts: 1658
UCSD
Gender: male
Re: Tipdate function
Reply #13 - Mar 28th, 2007 at 6:28pm
 
I have finally added the multiple rate (relaxed) molecular clock to today's build in the form of two new files under the Molecular Clock rubrik of Standard Analyses - GADatedClock.bf, GADatedClockProcessor.bf. In a nutshell, this method uses a GA to allocate multiple rates to tree branches which still enforcing a dated molecular clock. This is in unpublished (yet) method, but I will be writing some documentation on it shortly.

Cheers,
Sergei
Back to top
 

Associate Professor
Division of Infectious Diseases
Division of Biomedical Informatics
School of Medicine
University of California San Diego
WWW WWW  
IP Logged
 
Sergei
YaBB Administrator
*****
Offline


Datamonkeys are forever...

Posts: 1658
UCSD
Gender: male
Re: Tipdate function
Reply #14 - Mar 28th, 2007 at 6:33pm
 
Dear Jackson,

One of the sequences names must not be recognized.
If you download today's version of HyPhy, DatedTipsMolecularClock.bf will report which sequence failed to match the expected format - you should be able to debug it from there.

Cheers,
Sergei
Back to top
 

Associate Professor
Division of Infectious Diseases
Division of Biomedical Informatics
School of Medicine
University of California San Diego
WWW WWW  
IP Logged
 
Pages: 1 2