From owner-mol-evol@hgmp.mrc.ac.uk Mon May 1 15:56:23 2000 Return-Path: Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 110) id B94C417A61; Mon, 1 May 2000 15:56:21 +0100 (BST) Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 1) id 1376C17A59; Mon, 1 May 2000 15:56:19 +0100 (BST) To: mol-evol@net.bio.net Newsgroups: bionet.molbio.evolution Date: 29 Apr 2000 16:20:27 +0100 From: "Vijay Aswani, Ph.D." Subject: Parametric bootstrapping Message-Id: <20000501145619.1376C17A59@mercury.hgmp.mrc.ac.uk> Sender: owner-mol-evol@hgmp.mrc.ac.uk Precedence: bulk Hi everyone, I have a dataset of 12 taxa and ~4,000-odd bases that I am trying to analyze. Specifically, I am trying to test the monophyly of a sub-group of taxa on the tree. ModelTest selects GTR+I+G as the best model for my data under Maximum-Likelihood. I want to do the Log-Likelihood Ratio test on this hypothesis (monophyly of a sub-group within the tree). To test the significance I thought I would do parametric bootstrapping. Can anyone please show me how I go about this, step by step? I am thinking I would go about it as follows: 1. Generate simulated data sets using Seq-Gen (how many are appropriate?) Do I generate two groups of datasets - one with the best tree and the other with the best tree with the sub-group constrained as monophyletic?) 2. I guess I would then have to compute the likelihood scores for each of the simulated datasets. Assuming I did a 100, is there any automated way of doing this? Or do I have to open each dataset in PAUP*, load the appropriate tree, calculate likelihood scores and append them to a file? 3. What do I do next? Do I make a matrix doing substractions of every combination of likelihood values from the best tree set with those from the constrained tree set? Is there any program to do this? Is this what generates the distribution of delta values to which I would compare the delta obtained from the real data set? 4. Am I on the right track?! Has anyone out there done parametric bootstrapping to test monophyly by the method of Huelsenbeck et al (1996, 1997)? Can you share the procedural details of how you did this? Thanks in anticipation, to all who reply. Sincerely, Vijay /_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/ Vijay Aswani, Ph.D. Smithsonian Tropical Research Institute, Unit 0948, APO AA 34002-0948 U.S.A. Tel (in Panama): 507-212-8824 (work), 236-3243 (home) Fax: 507-228-0516 Email: vaswani@sinfo.net or aswaniv@naos.si.edu Homepage: http://aswani.freehosting.net /_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/ --- From owner-mol-evol@hgmp.mrc.ac.uk Mon May 1 15:57:14 2000 Return-Path: Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 110) id 1328217A63; Mon, 1 May 2000 15:57:10 +0100 (BST) Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 1) id A370017A61; Mon, 1 May 2000 15:57:06 +0100 (BST) To: mol-evol@net.bio.net Newsgroups: bionet.molbio.evolution Date: 1 May 2000 14:30:32 +0100 From: learn@valis.microbiol.washington.edu Subject: Charter for bionet.molbio.evolution Message-Id: <20000501145706.A370017A61@mercury.hgmp.mrc.ac.uk> Sender: owner-mol-evol@hgmp.mrc.ac.uk Precedence: bulk Information for MOLECULAR-EVOLUTION/bionet.molbio.evolution (moderated) USENET newsgroup name: bionet.molbio.evolution Status: Moderated One line Description: Discussions about research in molecular evolution. Moderation address: mol-evol-moderator@net.bio.net (mol-evol-moderator@net.bio.net is an alias for mol-evol-moderator@net.bio.net) Moderators: Jerry Learn James McInerney Mailing list name: MOLECULAR-EVOLUTION E-mail addresses: mol-evol@net.bio.net mol-evol@daresbury.ac.uk Newsgroup charter: Bionet.molbio.evolution is a forum for scientific discussions and a source of information for the community of scientists interested in the study of the processes of how DNA, RNA, proteins and organisms have evolved at a molecular level. Functions of the newsgroup: The newsgroup will facilitate rapid communication among research scientists, educators and other individuals interested in the study of molecular evolution. This includes, but is not limited to discussions of molecular phylogenies, methods of analysis, pointers towards sources of information, mechanisms of molecular evolution and software discussions. Finally, the newsgroup will serve as a bulletin board for announcements of meetings, conferences, job opportunities, and funding sources of interest to scientists studying molecular evolution. Subscribers are welcome from universities or any academic institutions, government agencies, hospitals and other medical institutions, and industrial or commercial organizations. Contributions within the functions outlined above are encouraged. Moderation Policy: Mass-posted commercial messages, chain letters, and similar postings not germane to molecular evolution will be deleted without comment. The newsgroup will not be used to facilitate discussions of a creationist nature. Inappropriate messages posted in good faith will be returned to the sender. Messages not strictly within the charter but likely to be of interest to many subscribers will be accepted. Appropriate messages are requests for information, announcements, or point-and-counterpoint on scientific issues, given and taken in a collegial spirit. Messages with obscenities, or messages that rely heavily on ad hominem arguments, may be rejected. Jerry Learn Dept. of Microbiology University of Washington Seattle, Washington USA email: learn@u.washington.edu James McInerney Dept. of Biology, National University of Ireland, Maynooth, Co. Kildare, Ireland email james.o.mcinerney@may.ie --- From owner-mol-evol@hgmp.mrc.ac.uk Tue May 2 16:33:46 2000 Return-Path: Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 110) id 0C25617B3A; Tue, 2 May 2000 16:33:44 +0100 (BST) Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 1) id 85F0117B1D; Tue, 2 May 2000 16:33:42 +0100 (BST) To: mol-evol@net.bio.net Newsgroups: bionet.molbio.evolution Date: Mon, 01 May 2000 19:24:45 -0400 From: John Huelsenbeck Subject: Re: Parametric bootstrapping Message-Id: <20000502153342.85F0117B1D@mercury.hgmp.mrc.ac.uk> Sender: owner-mol-evol@hgmp.mrc.ac.uk Precedence: bulk aswaniv@naos.si.edu Dear Vijay: I saw your post on bionet.molbio.evolution. I think I can answer your question. You calculate the log likelihood ratio test statistic for your original data. The test statistic is D = lnL0 - lnL1, where lnL0 is the log likelihood under the constraint that the subgroup of interest is monophyletic and lnL1 is the likelihood without the constraint. You want to generate the probability (null) distribution of D; this distribution is generated under the assumption that the subgroup really is monophyletic. How do you generate the null distibution? 1. Find the maximum likelihood tree, branch lengths, and substitution parameters under the null model (i.e., that the group is monophyletic). 2. Simulate 100 (or more) data sets using the tree from step #1. You are assured that the data were generated under the null model (i.e., you know for a fact that each simulated data set really did have the subgroup of interest monophyletic). 3. For each simulated data set, calculate lnL0 and lnL1. Find D for each simulated data set. 4. You will then have 100 (or more) D's. The frequency histogram of the D's calculated from the simulation represent an approximation of the null distribution. 5. If your observed D is greater than 95% of the simulated D's, then reject the null hypothesis (that the group is monophyletic) at the 5% level. You should be able to use SeqGen for step 2. I also have a program (available at http://brahms.biology.rochester.edu) that will simulate data under the HKY85+Gamma model of DNA substitution. The most time consuming part of the procedure is step # 3. You can automate this step as follows: Inbed all 100 simulated data sets into a single file. Your file will look something like this: #NEXUS begin data; dimensions ntax=x nchar=y; format datatype=dna; matrix Taxon 1 AACGT... Taxon 2 AACGG... etc... ; end; begin paup; [PAUP COMMANDS HERE] end; begin data; dimensions ntax=x nchar=y; format datatype=dna; matrix Taxon 1 AACGT... Taxon 2 AACGG... etc... ; end; begin paup; [PAUP COMMANDS HERE] end; begin data; dimensions ntax=x nchar=y; format datatype=dna; matrix Taxon 1 AACGT... Taxon 2 AACGG... etc... ; end; begin paup; [PAUP COMMANDS HERE] end; begin data; dimensions ntax=x nchar=y; format datatype=dna; matrix Taxon 1 AACGT... Taxon 2 AACGG... etc... ; end; begin paup; [PAUP COMMANDS HERE] end; begin data; dimensions ntax=x nchar=y; format datatype=dna; matrix Taxon 1 AACGT... Taxon 2 AACGG... etc... ; end; begin paup; [PAUP COMMANDS HERE] end; begin data; dimensions ntax=x nchar=y; format datatype=dna; matrix Taxon 1 AACGT... Taxon 2 AACGG... etc... ; end; begin paup; [PAUP COMMANDS HERE] end; (This is obviously meant to work with PAUP*). Search and replace all of the "[PAUP COMMANDS HERE]" with set autoclose=yes warnreset=no increase=auto; constraints my_constraint = ((1,2,3[whatever)); set criterion=parsimony; hsearch enforce=yes; set criterion=likelihood; lset nst=6 rmatrix=est basefreq=est rates=gamma shape=est pinvar=est; lscores 1; lset rmatrix=prev basefreq=prev shape=prev pinvar=prev; hsearch start=1 enforce=yes; lset nst=6 rmatrix=est basefreq=est rates=gamma shape=est pinvar=est; lscores 1; lset rmatrix=prev basefreq=prev shape=prev pinvar=prev; hsearch start=1 enforce=yes; set criterion=parsimony; hsearch enforce=no; set criterion=likelihood; lset nst=6 rmatrix=est basefreq=est rates=gamma shape=est pinvar=est; lscores 1; lset rmatrix=prev basefreq=prev shape=prev pinvar=prev; hsearch start=1 enforce=no; lset nst=6 rmatrix=est basefreq=est rates=gamma shape=est pinvar=est; lscores 1; lset rmatrix=prev basefreq=prev shape=prev pinvar=prev; hsearch start=1 enforce=no; This will perform two quick (quick for maximum likelihood, that is) searches, one under the constraint that the group is monophyletic and the other unconstrained. Make certain that on the first line of the the first paup block encountered, you put in the line log start file=myfile; and that the last line of the last paup block has log stop; That way, all of the results of the analysis will be output to a log file which you can read at your leisure. I don't have PAUP* on my computer at home, so the above paup block was from memory. Make a small test file to check it. On another note, you may also want to try the program BAMBE. BAMBE is written by Don Simon and Bret Larget (I don't know the URL, but you can reach Larget's page through the links page at my web site). The program performs Bayesian estimation of phylogeny. The program returns the posterior probabilities of trees (the posterior probability of a tree is the probability of the tree conditioned on the observed DNA sequences). The probability that the subgroup of interest is monophyletic is simply the sum of the posterior probabilities of trees having this group. This analysis will take a small fraction of the time that the LRT will take. Good Luck. John Huelsenbeck In article <8ek5ui$av5$1@mercury.hgmp.mrc.ac.uk>, "Vijay Aswani, Ph.D." wrote: >Hi everyone, > >I have a dataset of 12 taxa and ~4,000-odd bases that I am trying to >analyze. Specifically, I am trying to test the monophyly of a sub-group of >taxa on the tree. ModelTest selects GTR+I+G as the best model for my data >under Maximum-Likelihood. I want to do the Log-Likelihood Ratio test on this >hypothesis (monophyly of a sub-group within the tree). To test the >significance I thought I would do parametric bootstrapping. Can anyone >please show me how I go about this, step by step? > >I am thinking I would go about it as follows: > >1. Generate simulated data sets using Seq-Gen (how many are appropriate?) Do >I generate two groups of datasets - one with the best tree and the other >with the best tree with the sub-group constrained as monophyletic?) >2. I guess I would then have to compute the likelihood scores for each of >the simulated datasets. Assuming I did a 100, is there any automated way of >doing this? Or do I have to open each dataset in PAUP*, load the appropriate >tree, calculate likelihood scores and append them to a file? >3. What do I do next? Do I make a matrix doing substractions of every >combination of likelihood values from the best tree set with those from the >constrained tree set? Is there any program to do this? Is this what >generates the distribution of delta values to which I would compare the >delta obtained from the real data set? >4. Am I on the right track?! > >Has anyone out there done parametric bootstrapping to test monophyly by the >method of Huelsenbeck et al (1996, 1997)? Can you share the procedural >details of how you did this? > >Thanks in anticipation, to all who reply. > >Sincerely, > >Vijay > >/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/ >Vijay Aswani, Ph.D. >Smithsonian Tropical Research Institute, >Unit 0948, >APO AA 34002-0948 >U.S.A. >Tel (in Panama): 507-212-8824 (work), 236-3243 (home) >Fax: 507-228-0516 >Email: vaswani@sinfo.net or aswaniv@naos.si.edu >Homepage: http://aswani.freehosting.net >/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/ > > >--- From owner-mol-evol@hgmp.mrc.ac.uk Tue May 2 16:35:18 2000 Return-Path: Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 110) id 12C1417B43; Tue, 2 May 2000 16:35:16 +0100 (BST) Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 1) id 4BC8E17B3A; Tue, 2 May 2000 16:35:14 +0100 (BST) To: mol-evol@net.bio.net Newsgroups: bionet.molbio.evolution Date: 2 May 2000 11:51:00 +0100 From: Andrew Rambaut Subject: Re: Parametric bootstrapping Message-Id: <20000502153514.4BC8E17B3A@mercury.hgmp.mrc.ac.uk> Sender: owner-mol-evol@hgmp.mrc.ac.uk Precedence: bulk [[ This message was both posted and mailed: see the "To," "Cc," and "Newsgroups" headers for details. ]] In article <8ek5ui$av5$1@mercury.hgmp.mrc.ac.uk>, Vijay Aswani, Ph.D. wrote: > significance I thought I would do parametric bootstrapping. Can anyone > please show me how I go about this, step by step? Nick Goldman has a web page that describes the procedure: http://www.zoo.cam.ac.uk/zoostaff/goldman/index.html Click on the link that says "The Kishino-Hasegawa test of phylogenies is seriously biased: more information here...". > 1. Generate simulated data sets using Seq-Gen (how many are > appropriate?) Do Perhaps 200? Depends on how border-line the test statistic is. The nice thing about parametric bootstrapping is that if you have more than one machine available you can divide the task between them. > I generate two groups of datasets - one with the best tree and the other > with the best tree with the sub-group constrained as monophyletic?) No. You only generate the data on the NULL hypothesis. That is if your ML tree does not exhibit monophyly, then your NULL hypothesis is the best tree which does. The question you are asking is, "If the truth is that the group is monophyletic, is it likely that I got the non-monophyly result due to random error?" So you simulate on the NULL hypothesis and for each you perform exactly the same analysis that you did for your real data. > 2. I guess I would then have to compute the likelihood scores for each > of > the simulated datasets. Assuming I did a 100, is there any automated > way of > doing this? Or do I have to open each dataset in PAUP*, load the > appropriate > tree, calculate likelihood scores and append them to a file? Design a PAUP block with your PAUP commands for finding the ML trees under the monophyly constraints and without. Use the program Phy2Nex which is included in the Seq-Gen package to create a NEXUS file with your PAUP block inserted after each replicate dataset. Run this through PAUP. The best way of writing the likelihoods to a file is to (after the tree search) use the command: LSCORE 1 \ FILE=NULL.likelihoods APPEND=YES; This writes the likelihood of the tree to a file called NULL.likelihoods, appending each to the end of the file. > 3. What do I do next? Do I make a matrix doing substractions of every > combination of likelihood values from the best tree set with those from No you take the difference between the log likelihood for the monophyly constrained tree and the unconstrained tree for each simulated dataset. You then compare the difference in log likelihood you got for your real data. > constrained tree set? Is there any program to do this? Is this what Excel works well. Sort the simulated deltas and see what percentile your real delta falls at. > 4. Am I on the right track?! Nearly. Andrew --- From owner-mol-evol@hgmp.mrc.ac.uk Tue May 2 16:35:40 2000 Return-Path: Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 110) id C5AEB17B45; Tue, 2 May 2000 16:35:38 +0100 (BST) Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 1) id EDF0017B43; Tue, 2 May 2000 16:35:35 +0100 (BST) To: mol-evol@net.bio.net Newsgroups: bionet.molbio.evolution Date: 2 May 2000 12:52:47 +0100 From: Marcella Attimonelli Subject: median network software Message-Id: <20000502153535.EDF0017B43@mercury.hgmp.mrc.ac.uk> Sender: owner-mol-evol@hgmp.mrc.ac.uk Precedence: bulk I should be interested to know where it is possible to access to the software allowing the construction of the median network in order to classify haplotypes from nucleotide sequences of different subjects Thanks in advance marcella attimonelli ------------------------------------------------------------------- dr Marcella Attimonelli Dipartimento di Biochimica e Biologia Molecolare Via E.Orabona 4 - 70126 Bari - Italy Tel 39 080 5482130 FAX 39 080 5484467 e-mail marcella@area.ba.cnr.it Responsible of the Italian EMBnet node at Area di Ricerca del CNR - BARI, Italy Via Amendola 166/5 - 70126 Bari, Italy --- From owner-mol-evol@hgmp.mrc.ac.uk Thu May 4 20:00:50 2000 Return-Path: Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 110) id 2491F17B6A; Thu, 4 May 2000 20:00:49 +0100 (BST) Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 1) id 7884017B1C; Thu, 4 May 2000 20:00:47 +0100 (BST) To: mol-evol@net.bio.net Newsgroups: bionet.molbio.evolution Date: 4 May 2000 19:58:49 +0100 From: Vijay Aswani Subject: Parametric bootstrapping Message-Id: <20000504190047.7884017B1C@mercury.hgmp.mrc.ac.uk> Sender: owner-mol-evol@hgmp.mrc.ac.uk Precedence: bulk Hi all, First of all, I would like to thank all those who responded to my earlier posting asking for information on parametric bootstrapping. They were Chris Conroy, Andrew Rambaut, John Huelsenbeck, Thomas Buckley and Jack Sullivan. A number of those who responded pointed me to the web site of Nick Goldman who had a manuscript and some very helpful procedural tips on the process. I am writing to report that I have carried out the process and ... have some questions. First of all, the procedure seems to call for simulating datasets (I used a 100 replications) using the null hypothesis tree and estimated likelihood parameters of that tree. The null hypothesis tree, as I understand it, is not the best ML tree but the hypothetical topology I wish to test (eg. monophyly of a particular group). Once I have the 100 datasets, I then calculate the ML score of the best tree (by heuristic search) and the ML score of the hypothesis tree. I subtract: Lnull - LML and plot the distribution of these 100 values. I ran the test twice with 2 different hypotheses. I therefore simulated 100 datasets in each case. I noticed that in both of these results, the difference in ML values between the ML score of the best tree and the null hypothesis tree was very small. (As it turns out, in both cases, the hypotheses were rejected because of a larger difference between the real best ML score and null hypothesis ML score). This brings me to my question: isn't using the hypothesis tree's topology and ML parameters to build the 100 datasets and then computing the best tree in each dataset a bit circular. Wouldn't the best tree in each case be the same tree whose topology and ML parameters were used to create the data sets in the first place? Perhaps the reason why L null and L ML differ so little is that the dataset was created from the parameters of the null tree. If this is true, then the range of L ml - L null would be very small (since they would be almost the same) and almost every hypothesis tested would be rejected. I would appreciate any thoughts on this ... thanks, Vijay ________________________________________________ Vijay Aswani, Ph.D. Smithsonian Tropical Research Institute, Unit 0948 APO AA 34002-0948 U.S.A. Tel: +507-212-8824 Fax: +507-212-8790 Email: aswaniv@naos.si.edu or vaswani@sinfo.net ************************************************ --- From owner-mol-evol@hgmp.mrc.ac.uk Thu May 4 23:47:36 2000 Return-Path: Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 110) id 93C5517AB6; Thu, 4 May 2000 23:47:34 +0100 (BST) Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 1) id 5CFC417A7A; Thu, 4 May 2000 23:47:31 +0100 (BST) To: mol-evol@net.bio.net Newsgroups: bionet.molbio.evolution Date: 4 May 2000 20:52:55 +0100 From: Thomas Buckley Subject: RE: Parametric bootstrapping Message-Id: <20000504224731.5CFC417A7A@mercury.hgmp.mrc.ac.uk> Sender: owner-mol-evol@hgmp.mrc.ac.uk Precedence: bulk Hi Vijay, The parametric bootstrap test is powerful, in the statistical sense. However, it may become too liberal if the model you use to generate the sequences is too simple, relative to the "true model". So, we can attempt to minimize this potential error by using a complex and presumably realistic substitution model. The effect of inappropriate model assumptions has been shown in simulations (Huelsenbeck, 1996?), and in only one empirical study that I know of (Sullivan and Swofford, 1997). In my experience the null distribution is very strongly effected by the model used, and so the power of the test changes markedly with the substitution model used. So, I think you raise a good point that requires further study. Thomas -- Thomas Buckley Zoology Department Duke University Durham, NC 27708-0325 USA E-mail: tbuckley@duke.edu Phone: 919-660-7431 Fax: 919-684-6168 -----Original Message----- From: owner-mol-evol@hgmp.mrc.ac.uk [mailto:owner-mol-evol@hgmp.mrc.ac.uk] On Behalf Of Vijay Aswani Sent: Thursday, May 04, 2000 2:59 PM To: mol-evol@net.bio.net Subject: Parametric bootstrapping Hi all, First of all, I would like to thank all those who responded to my earlier posting asking for information on parametric bootstrapping. They were Chris Conroy, Andrew Rambaut, John Huelsenbeck, Thomas Buckley and Jack Sullivan. A number of those who responded pointed me to the web site of Nick Goldman who had a manuscript and some very helpful procedural tips on the process. I am writing to report that I have carried out the process and ... have some questions. First of all, the procedure seems to call for simulating datasets (I used a 100 replications) using the null hypothesis tree and estimated likelihood parameters of that tree. The null hypothesis tree, as I understand it, is not the best ML tree but the hypothetical topology I wish to test (eg. monophyly of a particular group). Once I have the 100 datasets, I then calculate the ML score of the best tree (by heuristic search) and the ML score of the hypothesis tree. I subtract: Lnull - LML and plot the distribution of these 100 values. I ran the test twice with 2 different hypotheses. I therefore simulated 100 datasets in each case. I noticed that in both of these results, the difference in ML values between the ML score of the best tree and the null hypothesis tree was very small. (As it turns out, in both cases, the hypotheses were rejected because of a larger difference between the real best ML score and null hypothesis ML score). This brings me to my question: isn't using the hypothesis tree's topology and ML parameters to build the 100 datasets and then computing the best tree in each dataset a bit circular. Wouldn't the best tree in each case be the same tree whose topology and ML parameters were used to create the data sets in the first place? Perhaps the reason why L null and L ML differ so little is that the dataset was created from the parameters of the null tree. If this is true, then the range of L ml - L null would be very small (since they would be almost the same) and almost every hypothesis tested would be rejected. I would appreciate any thoughts on this ... thanks, Vijay ________________________________________________ Vijay Aswani, Ph.D. Smithsonian Tropical Research Institute, Unit 0948 APO AA 34002-0948 U.S.A. Tel: +507-212-8824 Fax: +507-212-8790 Email: aswaniv@naos.si.edu or vaswani@sinfo.net ************************************************ --- --- From owner-mol-evol@hgmp.mrc.ac.uk Fri May 5 22:39:35 2000 Return-Path: Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 110) id 35FDD17AC3; Fri, 5 May 2000 22:39:34 +0100 (BST) Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 1) id DE8EE17A6F; Fri, 5 May 2000 22:39:32 +0100 (BST) To: mol-evol@net.bio.net Newsgroups: bionet.molbio.evolution Date: 5 May 2000 21:28:34 +0100 From: Jeff L.Blanchard Subject: Deadline for Madison Bioinformatics Courses Message-Id: <20000505213932.DE8EE17A6F@mercury.hgmp.mrc.ac.uk> Sender: owner-mol-evol@hgmp.mrc.ac.uk Precedence: bulk This summer the BioPharmaceutical Technology Center Institute in Madison, WI will be offering the following two bioinformatics-related courses. The application deadline for the courses is this Monday May 7th. After this date applicants will be accepted on a first come/ space available basis. However since the classes are limited to 16 participants additional space will be scarce, especially for "Techniques in Bioinformatics and Comparative Genomics". For more information see http://www.btci.org/courses/CoursesList/courses.htm or contact Jeff Blanchard at the address below. -------------------------------------------------------------------------------- Techniques in Bioinformatics and Comparative Genomics June 18 - June 23, 2000 Objectives and Goals: This five-day intensive computer laboratory course is designed to help participants construct a working library of bioinformatic tools and resources. The format will combine lectures and interactive problem-solving sessions with each participant working individually at a computer. An emphasis will be placed on analysis of practical problems posed by the participants. The flow of the course will move from traditional sequence analysis techniques to the opportunities afforded by the imminent flood of genomic information. Afternoon research seminars by the instructors will sample academic and private sector visions of bioinformatics and highlight creative approaches to utilizing genomic data. (Please note: this course is designed to provide an overview of the field; detailed training in specific packages will not be provided.) Instructors include: Jeff Blanchard, Ph.D., Research Scientist, National Center for Genome Resources Tim Burland, Ph.D., Vice President and General Manager, DNASTAR Barbara Butler, Ph.D., Bioinformatics Training and Education Group Leader, Genetics Computer Group Ross Overbeek, Ph.D., Senior Computer Scientist, Integrated Genomics Ann Palmenberg, Ph.D., Professor, Institute for Molecular Virology Department of Biochemistry,University of Wisconsin Michael Slater, Ph.D., Senior Scientist, Promega Corporation Jeff Thorne, Ph.D., Assistant Professor, Department of Statistics, North Carolina State University ---------------------------------------------------------------------- Database Design and Development for Genomics Research June 29 - July 1, 2000 Course Overview: Databases have been the quiet intermediary of molecular biology research and in their current state offer a wonderful example of using the web to share experimental results and knowledge. In recent years there has been a proliferation of individual and community oriented databases that contain DNA sequence, expression, structural, enzymatic, phenotypic, organismal and other types of data. Many scientists and private companies now face the challenge of integrating their data into these heterogeneous databases and/or synthesizing databases before they can finally get to the heart of a research question. This three-day "hands on" workshop will start with a session on "What are databases?" and move onto to database issues in molecular biology related to storing, integrating, visualizing, analyzing, synthesizing and presenting data. The workshop is geared for people in molecular biology lab groups and computer scientists getting into bioinformatics. The format will combine discussions with problem-solving sessions and each participant will work individually at a computer. Some specific topics include: An introduction to database systems and biological databases Using object-oriented principles to construct a relational database Developing and using gene expression databases Genomic Data Warehouse - Integrating data for Complex Analysis Integrating undisclosed/private and public data Development of Mouse Databases at the Jackson Laboratory A Distributed Sequence Annotation System Ontologies for Molecular Biology Databases Instructors include: Jeff Blanchard, Ph.D., Research Scientist, National Center for Genome Resources Carol Bult, Ph.D., Research Scientist, Mouse Genome Informatics Group, The Jackson Laboratory Allan Dickerman, Program Manager, National Center for Genome Resources Elizabeth Shoop, Ph.D., Research Associate, Computational Biology Center, University of Minnesota Lincoln Stein, Ph.D., Assistant Professor, Cold Spring Harbor Laboratory Jennifer Weller, Ph.D. Program Manager, National Center for Genome Resources There will probably be an additional instructor from the private sector. Matteo di Tommaso from Genetics Computer Group was originally scheduled to participate. However, his is in the process of moving to Celera and because of the time of his move will not be able to make it. ---------------------------------------------------------------------- The BioPharmaceutical Technology Center Institute (BTCI) is a not-for-profit organization operated exclusively for educational, scientific and cultural purposes. One goal of BTCI is to promote the exchange of scientific, educational and cultural information between industry, educators and the general public by providing facilities and resources to support conferences, seminars, classes, and electronic distribution of programs. _____________________________________ Jeffrey L. Blanchard Research Scientist National Center for Genome Resources 2935 Rodeo Park Drive East Santa Fe, NM 87505 tel: 505-995-4405 fax: 505-995-4432 http://www.ncgr.org/about/staff/jlb --- From owner-mol-evol@hgmp.mrc.ac.uk Mon May 8 17:44:52 2000 Return-Path: Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 110) id 9621C17AD5; Mon, 8 May 2000 17:44:51 +0100 (BST) Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 1) id 5819117A9D; Mon, 8 May 2000 17:44:50 +0100 (BST) To: mol-evol@net.bio.net Newsgroups: bionet.molbio.evolution Date: 8 May 2000 12:25:07 +0100 From: Nick Goldman Subject: Re: Parametric bootstrapping Message-Id: <20000508164450.5819117A9D@mercury.hgmp.mrc.ac.uk> Sender: owner-mol-evol@hgmp.mrc.ac.uk Precedence: bulk Vijay Aswani wrote: > > A number of those who responded pointed me to the web site of Nick Goldman > who had a manuscript and some very helpful procedural tips on the process. I should point out that the ms. in question (see http://www.zoo.cam.ac.uk/zoostaff/goldman/tests) is about problems that I and others contend exist with common uses of the Kishino-Hasegawa test of phylogenies. It is a lot more than simply a recipe for parametric bootstrapping, although it does include some parametric bootstrap examples. The ms. is "accepted pending minor revision"; these minor revisions are under way, and the revised version of the ms. should be complete this week and will be posted on the WWW site soon afterwards. Anyone is invited to e-mail me if they would like to be kept informed of progress. > This brings me to my question: isn't using the hypothesis tree's topology > and ML parameters to build the 100 datasets and then computing the best tree > in each dataset a bit circular. Wouldn't the best tree in each case be the > same tree whose topology and ML parameters were used to create the data sets > in the first place? Perhaps the reason why L null and L ML differ so little > is that the dataset was created from the parameters of the null tree. > > If this is true, then the range of L ml - L null would be very small (since > they would be almost the same) and almost every hypothesis tested would be > rejected. It is not circular. It is the strategy as used in most traditional statistics, i.e. "if the null hypothesis were true, what would be the distribution of my test statistic?". When the null hypothesis contains unknown parameters, you may have to estimate values for them in order to work out the distribution in question. So long as your method for working out the distribution in question allows for the fact that such parameters have been estimated (which is done by analyzing the simulated data in the same way that you analyzed the original data), the procedure is justified. Yes, often the distribution of "L_ml - L_null" is quite tight (small range). This will indeed often be because the ML tree for a simulated data set will be very similar to the null hypothesis tree on which the data were simulated. This simply reflects the situation under the assumption that the null hypothesis is true---and so it is appropriate to reject the null hypothesis in favour of the alternative hypothesis (which is that some other tree is correct). You have to think very carefully about what your hypotheses are before you test them! Nick Goldman ----------------------------------------------------------------------- Nick Goldman, Dept of Zoology, tel: +44-(0)1223-336649 Downing St, Cambridge CB2 3EJ, U.K. fax: +44-(0)1223-336679 N.Goldman@zoo.cam.ac.uk http://www.zoo.cam.ac.uk/zoostaff/goldman --- From owner-mol-evol@hgmp.mrc.ac.uk Tue May 9 00:10:28 2000 Return-Path: Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 110) id 1852617AA9; Tue, 9 May 2000 00:10:27 +0100 (BST) Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 1) id 8AAFB17A9F; Tue, 9 May 2000 00:10:25 +0100 (BST) To: mol-evol@net.bio.net Newsgroups: bionet.molbio.evolution Date: Mon, 08 May 2000 11:06:20 -0700 From: Guy Hoelzer Subject: Re: Parametric bootstrapping Message-Id: <20000508231025.8AAFB17A9F@mercury.hgmp.mrc.ac.uk> Sender: owner-mol-evol@hgmp.mrc.ac.uk Precedence: bulk In article <8f6qu1$d8t$1@mercury.hgmp.mrc.ac.uk>, Nick Goldman wrote: > Yes, often the distribution of "L_ml - L_null" is quite tight (small > range). This will indeed often be because the ML tree for a simulated > data set will be very similar to the null hypothesis tree on which the > data were simulated. This simply reflects the situation under the > assumption that the null hypothesis is true---and so it is appropriate > to reject the null hypothesis in favour of the alternative hypothesis > (which is that some other tree is correct). You have to think very > carefully about what your hypotheses are before you test them! This leads to a concern that I have about the use of LRTs and parametric bootstrapping. The limits of the null distribution are constrained (sometimes greatly) by the assumed evolutionary model, making it easier to reject the null by LRT. Therefore, the type-I error rate will be larger using this approach than if simpler models were assumed. The question is: does parametric bootstrapping lead to an unacceptibly high type-I error rate? Of course, this will only be a problem when the assumed evolutionary model is not accurate, which is always the case to some degree. For example, if you estimate a TI/TV ratio from your data, and assume the veracity of your estimate in your evolutionary model, it is probably the case that the TRUE TI/TV ration was somewhat different than the estimate for every branch in the TRUE tree. Therefore, the null distribution you create through repeated simulation is then guaranteed to differ from the universe of potential likelihoods that could have been explored during the evolution of your taxa. The realized variation in TI/TV ratios would surely broaden the TRUE null distribution, compared to the one estimated through simulations, leading to inflated type-I error rates in the analysis. I am curious if there is any evidence relating to this potential problem. -- Guy Hoelzer Department of Biology University of Nevada Reno Reno, NV 89557 From owner-mol-evol@hgmp.mrc.ac.uk Tue May 9 23:53:25 2000 Return-Path: Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 110) id 0E48417B3A; Tue, 9 May 2000 23:53:23 +0100 (BST) Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 1) id 9908D17AFC; Tue, 9 May 2000 23:53:22 +0100 (BST) To: mol-evol@net.bio.net Newsgroups: bionet.molbio.evolution Date: 9 May 2000 10:29:50 +0100 From: Nick Goldman Subject: Re: Parametric bootstrapping Message-Id: <20000509225322.9908D17AFC@mercury.hgmp.mrc.ac.uk> Sender: owner-mol-evol@hgmp.mrc.ac.uk Precedence: bulk Guy Hoelzer wrote: > > This leads to a concern that I have about the use of LRTs and parametric > bootstrapping. The limits of the null distribution are constrained > (sometimes greatly) by the assumed evolutionary model, making it easier to > reject the null by LRT. The statistics (LRTs and bootstraps) do what you ask them to. This comment is leading to a discussion of the usefulness of models, which is a different question to where this thread started. > Therefore, the type-I error rate will be larger > using this approach than if simpler models were assumed. The question is: > does parametric bootstrapping lead to an unacceptibly high type-I error > rate? Of course, this will only be a problem when the assumed > evolutionary model is not accurate, which is always the case to some > degree. For example, if you estimate a TI/TV ratio from your data, and > assume the veracity of your estimate in your evolutionary model, it is > probably the case that the TRUE TI/TV ration was somewhat different than > the estimate for every branch in the TRUE tree. Therefore, the null > distribution you create through repeated simulation is then guaranteed to > differ from the universe of potential likelihoods that could have been > explored during the evolution of your taxa. The realized variation in > TI/TV ratios would surely broaden the TRUE null distribution, compared to > the one estimated through simulations, leading to inflated type-I error > rates in the analysis. I am curious if there is any evidence relating to > this potential problem. I don't necessarily agree that the null distribution estimated by simulation is *guaranteed* to differ from the universe of distributions: for example, if the distribution is independent of the TI/TV ratio used when estimating it. Regular (asymptotic) statistics rely on the (asymptotic) independence of the null distribution and the unknown true values of parameters 'within' it. I do agree that the effect you describe *could* exist, particularly for small (whatever that means!) data sets. And no, I don't know of any studies that have investigated this potential effect in phylogenetic applications. Nick Goldman ----------------------------------------------------------------------- Nick Goldman, Dept of Zoology, tel: +44-(0)1223-336649 Downing St, Cambridge CB2 3EJ, U.K. fax: +44-(0)1223-336679 N.Goldman@zoo.cam.ac.uk http://www.zoo.cam.ac.uk/zoostaff/goldman --- From owner-mol-evol@hgmp.mrc.ac.uk Wed May 10 19:42:34 2000 Return-Path: Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 110) id 7FC1417ACC; Wed, 10 May 2000 19:42:32 +0100 (BST) Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 1) id 243C417A76; Wed, 10 May 2000 19:42:29 +0100 (BST) To: mol-evol@net.bio.net Newsgroups: bionet.molbio.evolution Date: 10 May 2000 18:49:40 +0100 From: Dimitris Stassinopoulos Subject: Workshop on Astrobiology at Alife7 Message-Id: <20000510184229.243C417A76@mercury.hgmp.mrc.ac.uk> Sender: owner-mol-evol@hgmp.mrc.ac.uk Precedence: bulk Dear Colleagues: We are organizing a workshop on Astrobiology at Alife7. The primary objective of this workshop is to highlight some of the new areas of research that have emerged form the new NASA Astrobiology initiative. We hope to have the final list of presentations from individuals working in some of these emerging areas of astrobiology soon. Since we also believe that there is great potential for a more permanent interactions between Astrobiology and the A-life community, and hope to foster such interaction through this workshop, we solicit a small number of shorter presentations of work that might benefit (or has already benefited) both communities. Please send an abstract of your proposed presentation to: stassi@groucho.arc.nasa.gov If your submission is accepted, be prepared to provide us with a short, camera-ready copy of your presentation. A more complete description of the workshop has already been posted in the Alife7 web-pages: http://alife7.alife.org/workshops.shtml#Astrobiology If you would like to learn more about Astrobiology, please visit the NASA web-sites on the subject: o http://astrobiology.arc.nasa.gov o http://cca.arc.nasa.gov o http://nai.arc.nasa.gov For details about camera-ready copy, please see: http://alife7.alife.org/workshops.shtml#SubmissionInstructions We are looking forward to receiving your contributions as well as your feedback and suggestions. >From the workshop organizers, Michael H. New (mike@groucho.arc.nasa.gov) Andrew Pohorille (pohorill@raphael.arc.nasa.gov) Dimitris Stassinopoulos (stassi@groucho.arc.nasa.gov) NASA Ames Research Center Exobiology Branch Mail Stop 239-4 Moffett Field, CA 94035 USA --- From owner-mol-evol@hgmp.mrc.ac.uk Sun May 14 17:51:12 2000 Return-Path: Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 110) id 1AC0417AD3; Sun, 14 May 2000 17:51:09 +0100 (BST) Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 1) id 051C117AD2; Sun, 14 May 2000 17:51:07 +0100 (BST) To: mol-evol@net.bio.net Newsgroups: bionet.molbio.evolution Date: 14 May 2000 14:49:31 +0100 From: Thomas Buckley Subject: HIV molecular phylogenetics Message-Id: <20000514165107.051C117AD2@mercury.hgmp.mrc.ac.uk> Sender: owner-mol-evol@hgmp.mrc.ac.uk Precedence: bulk Hi, I need an up to date reference for the latest on molecular phylogenetic relationships among HIV subtypes. Thanks, Thomas --- From owner-mol-evol@hgmp.mrc.ac.uk Tue May 16 16:53:00 2000 Return-Path: Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 110) id A040817AB1; Tue, 16 May 2000 16:52:57 +0100 (BST) Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 1) id 885C717A62; Tue, 16 May 2000 16:52:52 +0100 (BST) To: mol-evol@net.bio.net Newsgroups: bionet.molbio.evolution Date: 16 May 2000 16:48:43 +0100 From: Tim Littlewood Subject: 3 Yr Postdoctoral Position - NHM, London - please post Message-Id: <20000516155252.885C717A62@mercury.hgmp.mrc.ac.uk> Sender: owner-mol-evol@hgmp.mrc.ac.uk Precedence: bulk The enigma of the marine biodiversity focus in the Indo-West Pacific Ocean: competing models assessed by the molecular phylogeny of a species-rich molluscan clade Applicants are sought for the position of Postdoctoral Research Assistant on the above project, to be carried out in the new Molecular Biology Unit at the Natural History Museum, London. The project will start some time this year, to run for 3 years. Starting salary up to 25000 pounds. Please respond before 31 May 2000 with CV (including statement of research interests and starting availability) and e-mail addresses of two potential referees. This project aims to throw new light on two of the most longstanding and significant questions of evolutionary biology in the marine environment: - How does speciation occur in widely dispersed organisms? - How is the process related to the origin of the biodiversity focus of the Indo-West Pacific (IWP)? The study group, the littorinid genus Nodilittorina, is an eminently suitable model for the majority of marine animals contributing to the IWP diversity gradient; its 30 IWP species have a 4-week planktonic larval life and exhibit a range of both widespread and narrowly endemic geographical distributions. Samples of every known species of this genus are available, and distributions are accurately known. The application of molecular phylogenetics to IWP biogeography is only just beginning, and previous studies have been limited by restricted taxon sampling and poorly known distributions. It is therefore proposed: - For the first time, to generate a molecular phylogeny of a species-rich, monophyletic clade in the IWP. - To seek concordant patterns (e.g. sister-relations between Indian and Pacific Ocean sub-clades, and central or peripheral location of derived species) for comparison with the predictions of the four competing models of IWP biogeography (centre of origin, of accumulation, of survival, or of overlap). - Estimates of age from molecular divergence will distinguish between relatively recent (Pleistocene) and more ancient (Miocene or earlier) species radiations. - Further, the phylogeography of mitochondrial haplotypes will be examined from throughout the range of two widely distributed species in order to search for parallels between genetic and species diversity gradients, and between geographical occurrence of derived alleles and species. Such parallels will indicate how present-day determinants of gene flow (e.g. ocean currents and island isolation) contributed to speciation in the past. The ideal candidate will have skills in: a. standard molecular biology techniques +/- gene sequencing skills b. use of phylogenetic and population genetic software Facilities at the NHM are excellent for molecular systematic/phylogenetic work. For further details, contact: Dr David Reid or Dr Tim Littlewood Department of Zoology The Natural History Museum Cromwell Road London SW7 5BD United Kingdom http://www.nhm.ac.uk Tel: +44 (0) 20 7942 5051 or 5742 Fax: +44 (0) 20 7942 5054 E-mail: dgr@nhm.ac.uk or: t.littlewood@nhm.ac.uk _____________________________________ D.T.J. Littlewood Department of Zoology The Natural History Museum Cromwell Road, London SW7 5BD UK tel: +44 (0)207 942 5742 (office) 5008 (lab - you'll be lucky) fax: +44 (0)207 942 5151 _____________________________________ --- From owner-mol-evol@hgmp.mrc.ac.uk Thu May 18 00:07:37 2000 Return-Path: Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 110) id AEC7C17A81; Thu, 18 May 2000 00:07:36 +0100 (BST) Received: by mercury.hgmp.mrc.ac.uk (Postfix, from userid 1) id 8922E17A63; Thu, 18 May 2000 00:07:35 +0100 (BST) To: mol-evol@net.bio.net Newsgroups: bionet.molbio.evolution Date: Wed, 17 May 2000 15:44:49 +0200 From: Fernando Gonzalez Candelas Subject: First announcement. Workshop EVOLUTION, FROM MOLECULES TO ECOSYSTEMS Message-Id: <20000517230735.8922E17A63@mercury.hgmp.mrc.ac.uk> Sender: owner-mol-evol@hgmp.mrc.ac.uk Precedence: bulk Dear colleagues, We are pleased to invite you to attend and participate in the workshop entitled Evolution: From molecules to ecosystems that will be held 2-4 November 2000 in Valencia (Spain) as part of the activities commemorating the fifth centennial of the University of Valencia and the launching of the Cavanilles Institute for Biodiversity and Evolutionary Biology. The goal of the workshop will be to reflect on the present and foreseeable future of evolutionary theory in the age of genomics, development, and complex systems. Themes and Topics 1. Behavior and evolution 2. Coevolution 3. Comparative genomics 4. Evolution and development 5. Evolution at the organismic level and biodiversity 6. Evolutionary ecology 7. Evolution of life history traits 8. Evolution of parasites 9. Evolution of sex 10. Experimental evolution 11. Extinctions 12. Human evolution 13. Impact of Darwinism 14. Macroevolution 15. Molecular ecology 16. Molecular evolution 17. Origins and the roots of life 18. Phylogenetic analysis of major taxonomic groups 19. Phylogenetics and the comparative method 20. Speciation Speakers and invited participants include: M. Aguade (Spain), F. J. Ayala (USA), G. Barbujani (Italy), J. Bertranpetit (Spain), G.M. Burghardt (USA), A. Caballero (Spain), P. Calow (UK), S. Carroll (USA), L. Chao (USA), L. De Meester (Belgium), D.C. Dennett (USA), D. Erwin (USA), A. Fontdevila (Spain), P. Forterre (France), A. Garcia Bellido (Spain), C. Herrera (Spain), G.M. Hewitt (UK), D.M. Hillis (USA), P. Holland (UK), C.E. King (USA), S. Kresovich (USA), W. Lampert (Germany), A. Lazcano (Mexico), R.E. Lenski (USA), M. Lynch (USA), A. Mayer (Germany), R. Michod (USA), C. Mitter (USA), N. Moran (USA), H. Ochman (USA), T. Ohta (Japan), A.H. Patterson (USA), D. Pinero (Mexico), T.W. Snell (USA), S.J. Shettleworth (Canada), R. Thornhill (USA), M. Tibayrenc (France), R. Zardoya (Spain). Registration The workshop fee is 35000 ptas (210 euro), and includes registration, lunch, and refreshments served during coffee breaks at the venue. We encourage you to visit our web page for further information and on-line registration at: http://www.uv.es/cavanilles/evo2000 Attendance to the workshop will be limited to 180 participants. Selection, if necessary, will follow a first come, first served rule. Deadline for registration : 1 July 2000 Financial aid Support will be given where possible to a limited number of students attending the workshop. Two forms of financial aid will be offered, a fee waiver, and a fee waiver plus a fixed sum to cover travel expenses. Students must submit: 1) evidence that they are registered as a full-time student in an accredited program, such us an enrollment certificate or supporting letter from the university, department or academic advisor, 2) a short CV describing interests and achievements in the subject area of the workshop. It is important that applications be submitted as soon as possible in order to receive prompt consideration by the Organizing Committee. -- ******************************************************* Dr. Fernando Gonzalez Candelas Phone: (+34) 963 983 653 Instituto Cavanilles de Biodiversidad y Biologia Evolutiva Dept. de Genetica / Servei de Bioinformatica FAX : (+34) 963 983 670 Universitat de Valencia Apartado de Correos 22085 E-46071 Valencia SPAIN e-mail: Fernando.Gonzalez@uv.es *******************************************************