From C.Collins at MARLAB.AC.UK Thu Jun 4 17:57:33 2009 From: C.Collins at MARLAB.AC.UK (Catherine Collins) Date: Thu, 4 Jun 2009 17:57:33 +0100 Subject: [Sbforum-general] ML bootstrapping Message-ID: Hello. I would like to ask a question regarding bootstrapping maximum likelihood trees.? Currently we are generating trees with neighbour joining and maximum likelihood.? I would like to know if we can use bootstrap values generated by neighbour joining, using maximum likelihood settings, for the ML trees.?If not, is there an alternative way of bootstrapping the ML trees which?does not require a large computing resource? ? Thanks. Catherine Collins ############################################### Dr. Catherine Collins Molecular Genetics Marine Scotland - Science Scottish Government Marine Laboratory, PO Box 101, 375, Victoria Road, Aberdeen AB11 9DB Tel: +44 (0)1224 295683 Fax: +44 (0)1224 295511 e: collinsc at marlab.ac.uk w: http://www.scotland.gov.uk/marinescotland ############################################### From mark.blaxter at ed.ac.uk Mon Jun 15 15:30:31 2009 From: mark.blaxter at ed.ac.uk (Mark Blaxter) Date: Mon, 15 Jun 2009 15:30:31 +0100 Subject: [Sbforum-general] bootstrapping ML trees In-Reply-To: References: Message-ID: <7EF44766-CEB3-47B1-9D4C-B4167837F0C0@ed.ac.uk> Hi Catherine NO! you cant use the NJ bootstraps as a stand-in for the ML bootstraps. All you are doing is NJ analyses, not tree-space search in ML... If resource is an issue you should try Bayesian methods as in Mrbayes, which are both cpu effective, and generate meaningful searches of tree and model space in a very reasonable time space. Mark On 23 May 2009, at 19:44, Catherine Collins wrote: > > Hello. > I would like to ask a question regarding bootstrapping maximum > likelihood trees. > Currently we are generating trees with neighbour joining and maximum > likelihood. I would like to know if we can use bootstrap values > generated by neighbour joining, using maximum likelihood settings, > for the ML trees. If not, is there an alternative way of > bootstrapping the ML trees which does not require a large computing > resource? > > Thanks. > > Catherine Collins. > _______________________________________________ > Sbforum-general mailing list > Sbforum-general at sbforum.org > http://sbforum.org/mailman/listinfo/sbforum-general_sbforum.org Mark Blaxter mark.blaxter at ed.ac.uk ~ may all beings be happy ~ -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: not available URL: From jimp at compbio.dundee.ac.uk Mon Jun 15 16:20:01 2009 From: jimp at compbio.dundee.ac.uk (James Procter) Date: Mon, 15 Jun 2009 16:20:01 +0100 Subject: [Sbforum-general] bootstrapping ML trees In-Reply-To: <7EF44766-CEB3-47B1-9D4C-B4167837F0C0@ed.ac.uk> References: <7EF44766-CEB3-47B1-9D4C-B4167837F0C0@ed.ac.uk> Message-ID: <4A3666A1.8050709@compbio.dundee.ac.uk> Hi Catherine. Mark Blaxter wrote: > resource is an issue you should try Bayesian methods as in Mrbayes, > which are both cpu effective, and generate meaningful searches of tree > and model space in a very reasonable time space. Just to follow on from Mark's comment - you might find that TOPALi (see www.topali.org ) an easy way of running MrBayes analyses. best of luck Jim Procter. -- ------------------------------------------------------------------- J. B. Procter (ENFIN/VAMSAS) Barton Bioinformatics Research Group Phone/Fax:+44(0)1382 388734/345764 http://www.compbio.dundee.ac.uk The University of Dundee is a Scottish Registered Charity, No. SC015096. From db60 at st-andrews.ac.uk Wed Jun 17 17:18:15 2009 From: db60 at st-andrews.ac.uk (Daniel Barker) Date: Wed, 17 Jun 2009 17:18:15 +0100 Subject: [Sbforum-general] bootstrapping ML trees In-Reply-To: <4A378D22.9020503@st-andrews.ac.uk> References: <4A378D22.9020503@st-andrews.ac.uk> Message-ID: <4A391747.2060706@st-andrews.ac.uk> Re-sending - I originally sent to the 'bounces' address by mistake. Daniel I wrote: > Dear Catherine et al., > > A couple of additional suggestions. > > Some newer ML phylogeny software is very fast. It's still far slower > than NJ, but we've found it can allow bootstrapping where previously > this would have been difficult - > > PHYML: http://atgc.lirmm.fr/phyml > > Treefinder: http://www.treefinder.de > > Alternatively, in some software you can adjust the heuristic search for > each bootstrap replicate to be far more rapid than the default. E.g. > with PAUP (http://paup.csit.fsu.edu) one might try something like this - > 'In each replicate an initial neighbor-joining tree was used to estimate > the GTR model parameters, parameters were then fixed for that replicate, > the optimization was started with the neighbor-joining tree, and the > search used NNI branch swapping' > (http://dx.doi.org/10.1080/10635150600697465 Fig. 2). The crucial part > is not use of GTR (which depends on the input) but (a) beginning the > search for the ML tree with the NJ tree (which is often very similar to > the ML tree); (b) fixing GTR parameters at an early stage, rather than > estimating them jointly with branch lengths and topology; and (c) making > only slight rearrangements to the tree during the search (NNI). Such > 'shortcuts' do increase the risk that you won't find the true ML tree > topology and/or that your branch lengths will be wrong. So you might > want to reconstruct a single ML tree with a 'careful', slow search; and > then use rapid procedures such as this just to get bootstrap support > values. > > Best wishes, > > Daniel -- Daniel Barker http://bio.st-andrews.ac.uk/staff/db60.htm The University of St Andrews is a charity registered in Scotland : No SC013532 From Frank.Wright at scri.ac.uk Mon Jun 15 16:41:22 2009 From: Frank.Wright at scri.ac.uk (Frank Wright) Date: Mon, 15 Jun 2009 16:41:22 +0100 Subject: [Sbforum-general] bootstrapping ML trees (Catherine Collins) References: Message-ID: <50E41C3A73F46B4D876BD13F5E80264A02DAD64A@exchange3.sims.scri.sari.ac.uk> The recent revolution in statistical phylogenetics (from about year 2000 onwards) has resulted in the development of Maximum Likelihood (ML) and Bayesian tree methods that use increasingly realistic models of evolution. The recent software developments in ML have meant that we can avoid using Neighbor Joining methods unless we have 1000s of sequences. Recent approximate Maximum Likelihood phylogenetic software (for example PhyML, RaxML and Garli) can produce bootstrapped trees (say 100 bootstraps) in a few hours for 100-200 sequences. PhyML is slower and probably produced better tree estimates than RaxML. There is, however, a fast approximate PhyML bootstrap option which can speed things up. Some websites (e.g. CIPRES portal at http://www.phylo.org/sub_sections/portal/ and "Phylogeny.fr" at http://www.phylogeny.fr/) will provide access to cluster resources to run jobs. If you have 100-200 sequences, then MrBayes Bayesian tree are possible if you have the patience to wait a day or two as the CPU time does not increase as quickly with number of sequences. With MrBayes, you can use rDNA secondary structure information to produce better estimates of rDNA trees. If you load the MPI version of MrBayes on your (hopefully MPI-enabled) cluster then this should speed things up also. You could also carry out a quick approximate tree without bootstrapping and then use this tree to prune your dataset of identical or very-similar sequences. You could then submit the pruned dataset for bootstrap analysis using ML or Bayesian analysis. It is possible that careful "experimental design" will mean that you don't have to analysis all available data :-) Don't forget about model selection - choosing the appropriate nucleotide substitution model (or amino acid matrix) with appropriate rate heterogeneity model will improve ML and Bayesian tree estimation. Our program, TOPALi (www.topali.org), will compare models available in PhyML and MrBayes using AIC and BIC statistical criteria. The model selection jobs will run quickly (each model runs on a separate node on our cluster). TOPALi will also models to be chosen for each of the three codon positions if you have protein-coding DNA. These codon position models can then be used with MrBayes. TOPALi uses the "conventional" MODELTEST approach - however, for recent developments have a look at jMODELTEST (http://darwin.uvigo.es/software/jmodeltest.html ). TOPALi will also allow you to "prune" a tree (e.g. an approximate NJ tree) to remove sequences prior to running a PhyML or RaxML. TOPALi will also run PhyML, RaxML and MrBayes jobs but these three analyses have not yet been configured to utilise our cluster so will be a bit slow. You can run a "test" MrBayes job in TOPALi and then extract the MrBayes commands used by TOPALi for use with a MrBayes run launched using command-line MrBayes. RaxML also allows some modelling of codon positions for protein-coding DNA by allowing the GTR model to be estimated separately for each codon position. I hope this helps. Frank Wright BioSS frank at bioss.ac.uk >Date: Sat, 23 May 2009 19:44:57 +0100 >From: "Catherine Collins" > >I would like to ask a question regarding bootstrapping maximum likelihood trees. > >Currently we are generating trees with neighbour joining and maximum likelihood. I would >like to know if we can use bootstrap values generated by neighbour joining, using maximum >likelihood settings, for the ML trees. If not, is there an alternative way of bootstrapping the >ML trees which does not require a large computing resource? > >Thanks. >Catherine Collins. ______________________________________________________ SCRI, Invergowrie, Dundee, DD2 5DA. The Scottish Crop Research Institute is a charitable company limited by guarantee. Registered in Scotland No: SC 29367. Recognised by the Inland Revenue as a Scottish Charity No: SC 006662. DISCLAIMER: This email is from the Scottish Crop Research Institute, but the views expressed by the sender are not necessarily the views of SCRI and its subsidiaries. This email and any files transmitted with it are confidential to the intended recipient at the e-mail address to which it has been addressed. It may not be disclosed or used by any other than that addressee. If you are not the intended recipient you are requested to preserve this confidentiality and you must not use, disclose, copy, print or rely on this e-mail in any way. Please notify postmaster at scri.ac.uk quoting the name of the sender and delete the email from your system. Although SCRI has taken reasonable precautions to ensure no viruses are present in this email, neither the Institute nor the sender accepts any responsibility for any viruses, and it is your responsibility to scan the email and the attachments (if any). ______________________________________________________ -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/ms-tnef Size: 6747 bytes Desc: not available URL: From cjanssen at sbforum.org Thu Jun 18 12:51:05 2009 From: cjanssen at sbforum.org (Chris Janssen) Date: Thu, 18 Jun 2009 12:51:05 +0100 Subject: [Sbforum-general] IWPLS'09 Message-ID: <4A3A2A29.1090608@sbforum.org> CALL FOR PAPERS IWPLS'09 - International Workshop on Portals for Life Sciences Workshop details http://www.nesc.ac.uk/esi/events/1000/ Workshop location: e-Science Institute in Edinburgh Duration of the workshop: 14 - 15 September 2009 We invite the submission of papers related to various aspects of molecular and systems biology and portals. Suggested topics include, but are not limited to: - Life science data producing methods, such as mass spectrometry - Life science data processing methods, such as phylogenetic analyses - Modeling protein-carbohydrate recognition - Machine learning for computational immunomics - Molecular simulations for drug discovery - Portal technology and portal construction methods - Usage models and portal tools for life sciences - Security aspects of portals for life sciences - Usability studies of life sciences portals - Workflows and service composition for life sciences - Service discovery and ontologies for life sciences - Tools for managing grid and cloud portability - Demonstrations/ success stories Paper submission: 20 July 2009 Notification of acceptance: 14 August 2009 Camera ready papers: 24 August 2009 Best regards, Sandra Gesing University of T?bingen Wilhelm Schickard Institute for Computer Science Departement for Simulation of Biological Systems Sand 14 72076 T?bingen Germany Phone: +49-7071-2970456 Fax: +49-7071-295152 From Muriel.Mewissen at ed.ac.uk Tue Jun 23 10:33:53 2009 From: Muriel.Mewissen at ed.ac.uk (Muriel Mewissen) Date: Tue, 23 Jun 2009 10:33:53 +0100 Subject: [Sbforum-general] R users survey Message-ID: <9A08F04917AF40209E90324A09ACFF25@MMNTRACLAPTOP> The Division of Pathway Medicine and Edinburgh Parallel Computing Centre at the University of Edinburgh are currently working on a joint project to design a framework for using R in parallel called SPRINT. SPRINT is a Simple Parallel R INTerface and aims to allow researchers to easily access High Performance Computing to perform statistical analysis on post genomic data using R. We are now gathering input from R users on what specific functions create bottlenecks in their analysis. This information will be used to direct the developments of SPRINT, to ensure that key routines are optimised and that the functionality is immediately relevant to R users. It is your chance to tell us what R functions are specific bottleneck for your work. I encourage you to take part in our SPRINT online user requirements survey at https://www.survey.ed.ac.uk/sprint/. Taking part in the survey should only take around 15 minutes. Thank you for your valuable input. Please don't hesitate to contact me if you have any questions regarding SPRINT by emailing sprint at lists.ed.ac.uk. Kind regards, Muriel Mewissen --- Muriel Mewissen Division of Pathway Medicine University of Edinburgh Medical School The Chancellor's building 49 Little France Crescent Edinburgh, EH16 4SB -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336. From r.sinnott at nesc.gla.ac.uk Tue Jun 23 10:40:53 2009 From: r.sinnott at nesc.gla.ac.uk (Richard Sinnott) Date: Tue, 23 Jun 2009 10:40:53 +0100 Subject: [Sbforum-general] R users survey In-Reply-To: <9A08F04917AF40209E90324A09ACFF25@MMNTRACLAPTOP> Message-ID: <1EBF7BBBF6FFFC4A957DB8F74E1F765D023A2232@exchange-be4.centre.ad.gla.ac.uk> You may also want to have a look at multiR (http://www.ncess.ac.uk/tools/multir/) Cheers, Rich -----Original Message----- From: sbforum-general-bounces at sbforum.org [mailto:sbforum-general-bounces at sbforum.org] On Behalf Of Muriel Mewissen Sent: 23 June 2009 10:34 To: sbforum-general at sbforum.org Subject: [Sbforum-general] R users survey The Division of Pathway Medicine and Edinburgh Parallel Computing Centre at the University of Edinburgh are currently working on a joint project to design a framework for using R in parallel called SPRINT. SPRINT is a Simple Parallel R INTerface and aims to allow researchers to easily access High Performance Computing to perform statistical analysis on post genomic data using R. We are now gathering input from R users on what specific functions create bottlenecks in their analysis. This information will be used to direct the developments of SPRINT, to ensure that key routines are optimised and that the functionality is immediately relevant to R users. It is your chance to tell us what R functions are specific bottleneck for your work. I encourage you to take part in our SPRINT online user requirements survey at https://www.survey.ed.ac.uk/sprint/. Taking part in the survey should only take around 15 minutes. Thank you for your valuable input. Please don't hesitate to contact me if you have any questions regarding SPRINT by emailing sprint at lists.ed.ac.uk. Kind regards, Muriel Mewissen --- Muriel Mewissen Division of Pathway Medicine University of Edinburgh Medical School The Chancellor's building 49 Little France Crescent Edinburgh, EH16 4SB -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336. _______________________________________________ Sbforum-general mailing list Sbforum-general at sbforum.org http://sbforum.org/mailman/listinfo/sbforum-general_sbforum.org From sandra.borthwick at sbforum.org Fri Jun 26 18:07:01 2009 From: sandra.borthwick at sbforum.org (Sandra Borthwick) Date: Fri, 26 Jun 2009 18:07:01 +0100 Subject: [Sbforum-general] SBF Event : Comparative Genomics one-day International Conference 18th August 2009 Message-ID: <5292CE4F24EE449A859C8A04472B85C6@society.local> * * * * * * * * * * * * * * * Comparative Genomics One-day International Conference Date: 18th August 2009, 9:30 am Location: Bute Building, University of St Andrews * * * * * * * * * * * * * * * This is a one day event - A day of exciting talks on comparative genomics from an array of international speakers. Thousands of genome sequences are now available, and the amount of data on genomes continues to increase. This presents enormous opportunities to make discoveries of biological and medical interest, but also requires new approaches. Confirmed speakers: - Prof MARK BLAXTER, University of Edinburgh. "Comparative Nematode Genomics". - Prof GEOFF BARTON, University of Dundee. "A tale of three small RNAs". - Prof DANNIE DURAND, Carnegie Mellon University. "The Evolution of Multidomain Families". - Prof NEIL HALL, University of Liverpool. "Fast forward Genetics using second generation sequencing". - Dr MATTHEW HEGARTY, Aberystwyth University. "Genomic mergers: the consequences of hybridisation and polyploidy in Senecio". - Prof IAN KORF, University of California Davis Genome Center. "Intron mediated enhancement". - Dr DAVID MARTIN, University of Dundee. "Predicting Protein Function". - Dr ZEMIN NING, Wellcome Trust Sanger Institute. "Cancer genome assemblies and variation detection between normal and tumor human cells". - Prof ZIHENG YANG FRS, University College London. "Population genomics and human-chimpanzee speciation". - Prof ANDY WATERS, University of Glasgow, "Translation of Translation: Comparative genomics yields practical insights into sexual development in malaria parasites" More information can be found at http://www.sbforum.org/events.php?e_id=70 Registration details: Early registration is encouraged. Registration fees are ?40; students can register for a reduced fee of ?20. ** Please register on-line at: www.sbforum.org/events.php The event is followed by a networking buffet dinner. Please book early to help us gauge numbers for the catering http://www.sbforum.org/eregister.php?e_id=70 * * * * * * * * * * * * * * * We look forward to seeing you on Tuesday 18th August. Kind Regards Sandra Sandra Borthwick, Executive Assistant Scottish Bioinformatics Forum The Royal Society of Edinburgh 22-26 George Street Edinburgh EH2 2PQ Tel: +44 (0)131 240 2783 Fax: +44 (0)131 240 2786 email: sandra.borthwick at sbforum.org www.sbforum.org 10th International Conference on Systems Biology Edinburgh, UK. 10th-15th October 2010 http://www.icsb-2010.net/ The SBF is a project of the RSE Scotland Foundation is Scottish Charity No. SC024636 The information contained in this e-mail is confidential, intended for the above named individual/s and may be legally privileged. This message may contain personal views which are not the views of the Foundation/Forum, unless specifically stated www.rsescotlandfoundation.org.uk www.sbforum.org From emily.gardner at abdn.ac.uk Thu Jun 25 16:18:37 2009 From: emily.gardner at abdn.ac.uk (Gardner, Emily) Date: Thu, 25 Jun 2009 16:18:37 +0100 Subject: [Sbforum-general] Dynamics in Systems Biology, Aberdeen 2009 Message-ID: Good afternoon, I am assisting in organising conference, 'Dynamics in Systems Biology' that will take place at the University of Aberdeen in September. We would very much appreciate if you could announce this conference and possibly forward the attached conference poster to a list of potential attendees. Many thanks for your help in advance. Kind regards, Emily ---------------------------------------------------------------------- Emily Gardner Research Secretary School of Natural & Computing Sciences University of Aberdeen Room 165, Fraser Noble Building Aberdeen, AB24 3UE Tel 01224 272297 Fax 01224 272607 Email emily.gardner at abdn.ac.uk The University of Aberdeen is a charity registered in Scotland, No SC013683. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: DynSysBio-Poster.pdf Type: application/pdf Size: 897521 bytes Desc: DynSysBio-Poster.pdf URL: From Gillian.Sinclair at manchester.ac.uk Mon Jun 29 12:12:15 2009 From: Gillian.Sinclair at manchester.ac.uk (Gillian Sinclair) Date: Mon, 29 Jun 2009 12:12:15 +0100 Subject: [Sbforum-general] R users survey In-Reply-To: <1EBF7BBBF6FFFC4A957DB8F74E1F765D023A2232@exchange-be4.centre.ad.gla.ac.uk> References: <9A08F04917AF40209E90324A09ACFF25@MMNTRACLAPTOP> <1EBF7BBBF6FFFC4A957DB8F74E1F765D023A2232@exchange-be4.centre.ad.gla.ac.uk> Message-ID: <20090629121215609.00000002160@twiggeries> Hello Further to Richards email I'd also like to flag up that R is available on the UK National Grid Service (www.ngs.ac.uk) which is free at point of use for all UK academic researchers. A simple application form is all you have to complete to be able to use these resources. Further information can be found on the NGS website here - http://www.ngs.ac.uk/sites/ral/applications/analysis/R.html Best wishes Gillian Dr Gillian Sinclair Liaison Officer National Grid Service University of Manchester Research Computing Services Rm 73 Devonshire House Precinct Centre Oxford Road Manchester M13 9PL Email - gillian.sinclair at manchester.ac.uk Telephone - 0161 275 0673 Fax - 0161 275 6120 NGS Website - www.ngs.ac.uk Keep up to date with all the news from the NGS at www.jiscmail.ac.uk/NGS-NEWS The NGS blog can be found at http://www.nationalgridservice.blogspot.com/ The latest edition of the NGS newsletter can be found at - www.ngs.ac.uk/newsletter -----Original Message----- From: sbforum-general-bounces at sbforum.org [mailto:sbforum-general-bounces at sbforum.org] On Behalf Of Richard Sinnott Sent: 23 June 2009 10:41 To: Muriel Mewissen; sbforum-general at sbforum.org Subject: Re: [Sbforum-general] R users survey You may also want to have a look at multiR (http://www.ncess.ac.uk/tools/multir/) Cheers, Rich -----Original Message----- From: sbforum-general-bounces at sbforum.org [mailto:sbforum-general-bounces at sbforum.org] On Behalf Of Muriel Mewissen Sent: 23 June 2009 10:34 To: sbforum-general at sbforum.org Subject: [Sbforum-general] R users survey The Division of Pathway Medicine and Edinburgh Parallel Computing Centre at the University of Edinburgh are currently working on a joint project to design a framework for using R in parallel called SPRINT. SPRINT is a Simple Parallel R INTerface and aims to allow researchers to easily access High Performance Computing to perform statistical analysis on post genomic data using R. We are now gathering input from R users on what specific functions create bottlenecks in their analysis. This information will be used to direct the developments of SPRINT, to ensure that key routines are optimised and that the functionality is immediately relevant to R users. It is your chance to tell us what R functions are specific bottleneck for your work. I encourage you to take part in our SPRINT online user requirements survey at https://www.survey.ed.ac.uk/sprint/. Taking part in the survey should only take around 15 minutes. Thank you for your valuable input. Please don't hesitate to contact me if you have any questions regarding SPRINT by emailing sprint at lists.ed.ac.uk. Kind regards, Muriel Mewissen --- Muriel Mewissen Division of Pathway Medicine University of Edinburgh Medical School The Chancellor's building 49 Little France Crescent Edinburgh, EH16 4SB -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336. _______________________________________________ Sbforum-general mailing list Sbforum-general at sbforum.org http://sbforum.org/mailman/listinfo/sbforum-general_sbforum.org _______________________________________________ Sbforum-general mailing list Sbforum-general at sbforum.org http://sbforum.org/mailman/listinfo/sbforum-general_sbforum.org From sandra.borthwick at sbforum.org Tue Jun 30 07:56:00 2009 From: sandra.borthwick at sbforum.org (sandra.borthwick at sbforum.org) Date: Tue, 30 Jun 2009 07:56:00 +0100 Subject: [Sbforum-general] Out of office Message-ID: Thank you for your message. I am out of office until 8th July, and will reply to you as soon as possible after my return. Should you have urgent business, please email Dr. Chris Janssen at cjanssen at sbforum.org Regards Sandra