April 2016 - phenixbb - phenix-online.org

Re: [phenixbb] Data processing: exhaustive search?
by Kay Diederichs 17 Apr '16

17 Apr '16

Hi Kaushik, there is no need for, and no sense in, an exhaustive search of data processing parameter space, because its dimensionality is way too high to achieve anything useful. The programs MOSFLM and XDS have undergone decades of development, and their processing defaults are carefully chosen. It is rather the user of the programs who is in charge to correctly interpret the output (which requires reading of their documentation, and some experience), and to make meaningful adjustments to (usually very) few parameters. Speaking for XDS, I recommend to read XDSwiki articles (there are tutorials and example data sets, and lots of explanations) and a book chapter that I recently wrote (#131 at http://cms.uni-konstanz.de/strucbio/diederichs-group/publications ). If you use the generate_XDS.INP script from XDSwiki, and follow the paper and the "Optimization" article, you will get good data from XDS. (There are other scripts, like autoPROC, xia2, xdsme and so on, which usually also give good results but I don't know them well.) In particular, in case of XDS there is usually no need to fiddle around with "box size, spot separation, tolerance and other parameters". The fact that you seem to try and optimize Rmerge suggests to me that you are trading precision for accuracy. Too many people still do this, but it is a legacy of the past. "Trying a combination of better frames over poorer frames" is most likely _not_ going to improve your merged intensities - this is another unfortunate and very common misunderstanding (see references 113 and 130). I hope this helps you to get better data. Kay On 16.04.2016 21:18, phenixbb-request(a)phenix-online.org wrote: > Message: 6 > Date: Sun, 17 Apr 2016 00:48:31 +0530 > From: Kaushik Hatti <hskaushik(a)gmail.com> > To: PHENIX user mailing list <phenixbb(a)phenix-online.org> > Subject: [phenixbb] Data processing: exhaustive search? > Message-ID: > <CAGMFGb=B9s-j4yXmRx8_M7NPndf86puviiY9QMg=EAqaWr9thw(a)mail.gmail.com> > Content-Type: text/plain; charset="utf-8" > > Hello, > > I have a data set diffracted to 2.7A collected at 1.54 wavelength. The data > is not of great quality with very close spots, ellipsoid spots in certain > regions of the frames and spot overlaps. Also certain frames are poorer > than rest. The crystal could not be reproduced. > > The best R merge (overall) I have achieved so far is 17% with 85% > completeness and 3 I/SigI processed in P4 space group. I feel it's possible > to process it better by identifying right values for box size, spot > separation, tolerance and other parameters. I am also suspecting a higher > symmetry space group. I believe, trying a combination of better frames > over poorer ones with improve merging statistics. > > I have so far tried processing in iMosflm and XDS. Is there a tool which > could search exhaustively trying different values for parameters and > suggest the combination which provide best merging statistics? > > Sorry if this question is not relevant in this group. > Any pointers/suggestions would be gratefully helpful. > > Thanks in advance, > Regards, > Kaushik > Molecular Biophysics Unit, > Indian Institute of Science, > Bangalore, India. -- Kay Diederichs http://strucbio.biologie.uni-konstanz.de email: Kay.Diederichs(a)uni-konstanz.de Tel +49 7531 88 4049 Fax 3183 Fachbereich Biologie, Universität Konstanz, Box 647, D-78457 Konstanz

2 1

Data processing: exhaustive search?
by Kaushik Hatti 16 Apr '16

16 Apr '16

Hello, I have a data set diffracted to 2.7A collected at 1.54 wavelength. The data is not of great quality with very close spots, ellipsoid spots in certain regions of the frames and spot overlaps. Also certain frames are poorer than rest. The crystal could not be reproduced. The best R merge (overall) I have achieved so far is 17% with 85% completeness and 3 I/SigI processed in P4 space group. I feel it's possible to process it better by identifying right values for box size, spot separation, tolerance and other parameters. I am also suspecting a higher symmetry space group. I believe, trying a combination of better frames over poorer ones with improve merging statistics. I have so far tried processing in iMosflm and XDS. Is there a tool which could search exhaustively trying different values for parameters and suggest the combination which provide best merging statistics? Sorry if this question is not relevant in this group. Any pointers/suggestions would be gratefully helpful. Thanks in advance, Regards, Kaushik Molecular Biophysics Unit, Indian Institute of Science, Bangalore, India.

1 0

Re: [phenixbb] Hardward phenix.refine
by Kay Diederichs 13 Apr '16

13 Apr '16

Colin, check out the first option in the FAQ at https://www.phenix-online.org/documentation/faqs/refine.html A core i7 and a Xeon of the same generation are equally fast if their frequency is the same. The main differences are that - Xeon E3 supports ECC memory (most core i* don't) - Xeon E5 supports Dual-CPU (core i* is just one) - Xeon E7 supports multi-CPU and is very expensive - most Xeons don't have built-in graphics whereas most core i* do There are many other small differences but most of them irrelevant; inspect the Wikipedia pages! Xeons won't be faster with phenix.refine; a core i7 already has enough cores - to increase their usage you'll have to recompile as per the FAQ. HTH, Kay > > Message: 1 > Date: Mon, 11 Apr 2016 07:56:39 +0000 > From: Colin Levy <C.Levy(a)manchester.ac.uk> > To: "phenixbb(a)phenix-online.org" <phenixbb(a)phenix-online.org> > Subject: [phenixbb] Hardward phenix.refine > Message-ID: <773C4580-FA01-4246-86E9-2D73AFD98421(a)manchester.ac.uk> > Content-Type: text/plain; charset="us-ascii" > > Dear all, > > Having read the FAQ regarding hardware I have a few ideas but would greatly appreciate any input from the community or developers. > > Would anyone like to suggest the best configuration of hardware for running phenix.refine? I mainly operate in OSX but have no issue with switching operating system if I can get a faster setup. My current thoughts are either something running a 4GHz Quad core i7 or an alternative would be a 3.7GHz Quad-Core Intel Xeon E5 (iMac v Mac Pro) or equivalent Linux builds. > > I know that the i7 is reasonably fast as I have this setup in a different machine and phenix.refine runs reasonably quickly but as ever I would like it to be faster. Budget is not unlimited by any means but I would be interested in learning what an optimal setup might look like. > > Many thanks, > > Colin > > > Manchester > Protein > Structure > Facility > > Dr. Colin W. Levy > MIB G034 > Tel. 0161 275 5090 > Mob.07786 197 554 > c.levy(a)manchester.ac.uk<mailto:[email protected]> > > Manchester Institute of Biotechnology | University of Manchester | 3.020 Garside Building | 131 Princess Street | Manchester | M1 7DN

3 2

Re: [phenixbb] Phenix Xtriage label error
by Alex Lee 12 Apr '16

12 Apr '16

Hi Dorothee and Billy, Thank you for your information! I have both I, SIGI and IPR, SIGIPR labels in my unmerged MTZ (I checked this with ViewHKL). But it seems the Xtriage only offers choices to merged mtz. I tried to choose I,SIGI, Merged option (even though my mtz is unmerged), the Xtriage output shows "anomalous flag false" and I did not get any additional anomalous information from the report. On Tue, Apr 12, 2016 at 12:16 PM, Dorothee Liebschner <dcliebschner(a)lbl.gov> wrote: > Hi Alex, > > What data labels are in the file pointless_xxx.mtz? > You can check with > > phenix.mtz.dump pointless_xxx.mtz > The output should list the available labels. > (you should see the labels also in hklview) > > You cannot leave the data labels choice blank in xtriage, as you have two > data choices in your file, so you need to tell the program which ones to > use. > > Apart from xtriage, it is always helpful to check the anomalous signal > reported in various log files from your data processing steps, such as > aimless.log and also from mosflm (I am not familiar to mosflm but I guess > it outputs also some info about anomalous signal?). > > Best wishes, > > Dorothee > > > > > > On Tue, Apr 12, 2016 at 11:08 AM, Alex Lee <alexlee198609(a)gmail.com> > wrote: > >> Dear Phenixbb members, >> >> I have a data-set of a 8 kDa protein crystal around 2.5A resolution. The >> protein crystal was soaked in potassium iodine before collecting data using >> in-house beam (1.54A wavelength). As I expected some anomalous signal from >> the iodine ion from the crystal, after I use Imosflm to index, integrate >> and scale the data (space group P3). I got four output .mtz files: >> pointless_XXX.mtz; aimless_xxx.mtz; ctruncate_xxx.mtz; >> ctruncate_xxx-unique.mtz. After I check with viewHKL, I found the only >> unmerged mtz data is pointless_XXX.mtz, the other three mtz files are >> merged. >> >> The next step I tried to use Phenix Xtriage (Linux version 1.10.1) to >> check my mtz data for anomalous completeness, I thought in this step my mtz >> should be unmerged type to see the anomalous signal, so I chose >> "pointless_xxx.mtz" as Xtriage input, but for the data labels in the >> Xtriage GUI panel, I can only have two choices of "I, SIGI, Merged" and >> "IPR, SIGIPR, Merged", it seems I do not have a choice of an unmerged mtz >> label. I decide to leave this choice blank by choosing data labels"---". >> After I click "run", Xtriage gave an error "please select labels for input >> data". >> >> Any input on this issue? >> >> Thanks in advance. >> >> >> _______________________________________________ >> phenixbb mailing list >> phenixbb(a)phenix-online.org >> http://phenix-online.org/mailman/listinfo/phenixbb >> Unsubscribe: phenixbb-leave(a)phenix-online.org >> > >

3 2

Phenix Xtriage label error
by Alex Lee 12 Apr '16

12 Apr '16

Dear Phenixbb members, I have a data-set of a 8 kDa protein crystal around 2.5A resolution. The protein crystal was soaked in potassium iodine before collecting data using in-house beam (1.54A wavelength). As I expected some anomalous signal from the iodine ion from the crystal, after I use Imosflm to index, integrate and scale the data (space group P3). I got four output .mtz files: pointless_XXX.mtz; aimless_xxx.mtz; ctruncate_xxx.mtz; ctruncate_xxx-unique.mtz. After I check with viewHKL, I found the only unmerged mtz data is pointless_XXX.mtz, the other three mtz files are merged. The next step I tried to use Phenix Xtriage (Linux version 1.10.1) to check my mtz data for anomalous completeness, I thought in this step my mtz should be unmerged type to see the anomalous signal, so I chose "pointless_xxx.mtz" as Xtriage input, but for the data labels in the Xtriage GUI panel, I can only have two choices of "I, SIGI, Merged" and "IPR, SIGIPR, Merged", it seems I do not have a choice of an unmerged mtz label. I decide to leave this choice blank by choosing data labels"---". After I click "run", Xtriage gave an error "please select labels for input data". Any input on this issue? Thanks in advance.

3 4

Hardward phenix.refine
by Colin Levy 11 Apr '16

11 Apr '16

Dear all, Having read the FAQ regarding hardware I have a few ideas but would greatly appreciate any input from the community or developers. Would anyone like to suggest the best configuration of hardware for running phenix.refine? I mainly operate in OSX but have no issue with switching operating system if I can get a faster setup. My current thoughts are either something running a 4GHz Quad core i7 or an alternative would be a 3.7GHz Quad-Core Intel Xeon E5 (iMac v Mac Pro) or equivalent Linux builds. I know that the i7 is reasonably fast as I have this setup in a different machine and phenix.refine runs reasonably quickly but as ever I would like it to be faster. Budget is not unlimited by any means but I would be interested in learning what an optimal setup might look like. Many thanks, Colin Manchester Protein Structure Facility Dr. Colin W. Levy MIB G034 Tel. 0161 275 5090 Mob.07786 197 554 c.levy(a)manchester.ac.uk<mailto:[email protected]> Manchester Institute of Biotechnology | University of Manchester | 3.020 Garside Building | 131 Princess Street | Manchester | M1 7DN [cid:[email protected]]<https://www.facebook.com/MIB2006/?ref=aymt_homepage_panel>[cid:[email protected]]<https://plus.google.com/u/0/b/111834792186886067949/111834792186886067949/p…>[cid:[email protected]]<https://www.linkedin.com/company/manchester-institute-of-biotechnology?trk=…>[cid:[email protected]]<https://twitter.com/?lang=en-gb&lang=en-gb>[cid:[email protected]]<https://www.youtube.com/channel/UCCBV8zDxk4L9yOEcujvSFNw> <https://www.youtube.com/channel/UCCBV8zDxk4L9yOEcujvSFNw><https://plus.google.com/u/0/b/111834792186886067949/111834792186886067949/p…><https://plus.google.com/u/0/b/111834792186886067949/111834792186886067949/p…><https://plus.google.com/u/0/b/111834792186886067949/111834792186886067949/p…>

2 1

Call for targets for CASP12 - Critical Assessment of Protein Structure Prediction
by Torsten Schwede 08 Apr '16

08 Apr '16

Dear colleagues, The CASP experiment is looking for prediction targets: Novel folds, membrane proteins, protein complexes and multimers, comparative modeling targets. As many of you know, the CASP community experiments have been running once every two years since 1994, collecting information on soon to be solved structures from the experimental community, and passing on sequence data to the structure modeling community so that blind predictions of structure can be collected and assessed. Over that period CASP has seen enormous progress in the quality of modeled structures, but many problems remain, and further testing of prediction methods is imperative for advancing the field. CASP organizers collect target information for a three month season every two years, and in that period we can acquire around 100 targets, sufficient to evaluate the state of the art for most types of modeling. For those of you who have not provided targets to CASP before, the procedure is simple - there is a web page for submitting targets, and a very experienced staff to deal with any queries. We don't need the structure in advance of its release by the PDB, and if we are notified early enough (a minimum of three weeks before release, more is better) there need be no delay in structure release. We need all sorts of targets and in general everything that has low coverage by the templates (<70% of the sequence length) and relatively low sequence identity to the best template (<50% ) will be appreciated. In particular, we are interested in: 1. Novel folds and membrane protein targets. Last round we observed some interesting methods developments for contact prediction and ab initio modeling, and it is important to be able to decisively evaluate their effectiveness this CASP again. 2. Protein complexes and multimers. Since the last edition of CASP, the assessment of structure prediction methods has stronger emphasis on evaluating the accuracy of quaternary structure prediction. Both oligomeric targets and complexes are of interest for the experiment. 3. A diversity of comparative modeling targets. Cases where there is fairly high sequence identity (30-50%) between the target structure and an available template are valuable for testing the degree to which model accuracy can approach that of experiment, particularly in functionally critical regions. Cases with lower sequence identity to template, right down to undetectable, are valuable for testing the ability of the methods to detect remote homologs, to overcome challenging alignment difficulties, to make use of multiple templates, and to build regions of the structure not obviously available from a template. The time table is similar to previous CASPs: The prediction season opens at the beginning of May, and will run until the end of July. We are releasing targets continuously throughout that period, as evenly spaced as possible, aiming for about 100 targets altogether. Each target will be available for prediction for a period of three weeks. For cases where longer periods are possible, these targets may be used to test refinement and data-assisted methods. It is of course important that there not be any kind of public release of the experimental structure (including things like pictures on web pages or abstracts) until after the predictions for that target are closed. As many of you know, it's fairly simple, with just two things to bear in mind. First, because of the timing framework, there needs to be at least a month between the submission date and any release of the structure. Second, we ideally need the experimental co-ordinates by the beginning of August and definitely by the end of August, so that the predictions can be assessed. At that point, they can be kept confidential if necessary, though we would like to provide them to predictors of your structure at the beginning of November at the latest, so that can see how well they have done. Participants would also usually like to be able to show slides and discuss their predictions at the meeting at the beginning of December. So, if you are currently working on an interesting protein structure suitable as prediction target for CASP , we would be most grateful if you would inform us via the target entry page: http://www.predictioncenter.org/casp12/targets_submission.cgi After CASP prediction season is over we usually publish a paper on the experimentalists' insights into the most interesting targets (see below). 2015: Some of the most interesting CASP11 targets through the eyes of their authors. Kryshtafovych A, Moult J, Baslé A, Burgin A, Craig TK, Edwards RA, Fass D, Hartmann MD, Korycinski M, Lewis RJ, Lorimer D, Lupas AN, Newman J, Peat TS, Piepenbrink KH, Prahlad J, van Raaij MJ, Rohwer F, Segall AM, Seguritan V, Sundberg EJ, Singh AK, Wilson MA, Schwede T. Proteins. 2015 Oct 16. doi: 10.1002/prot.24942. [Epub ahead of print]. PMID: 26473983. 2014: Challenging the state of the art in protein structure prediction: Highlights of experimental target structures for the 10th Critical Assessment of Techniques for Protein Structure Prediction Experiment CASP10. Kryshtafovych A, Moult J, Bales P, Bazan JF, Biasini M, Burgin A, Chen C, Cochran FV, Craig TK, Das R, Fass D, Garcia-Doval C, Herzberg O, Lorimer D, Luecke H, Ma X, Nelson DC, van Raaij MJ, Rohwer F, Segall A, Seguritan V, Zeth K, Schwede T. Proteins. 2014 Feb;82 Suppl 2:26-42. doi: 10.1002/prot.24489. Erratum in: Proteins. 2015 Jun;83(6):1198. PMID: 24318984. 2011: Target highlights in CASP9: Experimental target structures for the critical assessment of techniques for protein structure prediction. Kryshtafovych A, Moult J, Bartual SG, Bazan JF, Berman H, Casteel DE, Christodoulou E, Everett JK, Hausmann J, Heidebrecht T, Hills T, Hui R, Hunt JF, Seetharaman J, Joachimiak A, Kennedy MA, Kim C, Lingel A, Michalska K, Montelione GT, Otero JM, Perrakis A, Pizarro JC, van Raaij MJ, Ramelot TA, Rousseau F, Tong L, Wernimont AK, Young J, Schwede T. Proteins. 2011;79 Suppl 10:6-20. doi: 10.1002/prot.23196. Epub 2011 Oct 21. PMID: 22020785. Thanks, CASP organizing committee: John Moult, University of Maryland, USA Krzysztof Fidelis, University of California, Davis, USA Andriy Kryshtafovych, University of California, Davis, USA Torsten Schwede, University of Basel, Switzerland Anna Tramontano, University of Rome, Italy Get in touch: casp(a)predictioncenter.org<mailto:[email protected]> More information: http://www.predictioncenter.org/casp12/index.cgi Submit a target: http://www.predictioncenter.org/casp12/targets_submission.cgi

1 0

Position available with Phaser team at University of Cambridge
by Randy Read 06 Apr '16

06 Apr '16

I’d like to draw your attention to an opening that we have for someone to join the team developing Phaser. The ideal person would have experience with C++ and/or Python, graphical interfaces, and knowledge of how macromolecular structures are solved and validated, but it would also be good to hear from people who only tick some of those boxes. Full details of the position and the application process are given on the jobs.ac.uk website: http://www.jobs.ac.uk/job/ANJ468/research-associate/. I’m happy to answer any informal enquiries about the position. Best wishes, Randy Read ----- Randy J. Read Department of Haematology, University of Cambridge Cambridge Institute for Medical Research Tel: +44 1223 336500 Wellcome Trust/MRC Building Fax: +44 1223 336827 Hills Road E-mail: rjr27(a)cam.ac.uk Cambridge CB2 0XY, U.K. www-structmed.cimr.cam.ac.uk

1 0

Modeller- homology modelling
by Shilpa Janarthanan 03 Apr '16

03 Apr '16

Hello, Has anyone encountered this error while running Modeller for doing Homology modelling "Sequence mismatch between alignment and pdb"? If yes, how did you solve it? I checked, there is no mismatch between sequence(both fasta and pdb) and alignment. Thank you. -- Shilpa J Bioinformatics Dept., SASTRA University.

1 0

Finding the transcript
by Shilpa Janarthanan 02 Apr '16

02 Apr '16

Hello, Can you tell me, on what basis should I find the exact transcript (mRNA) sequence for a given protein sequence, in NCBI -Gene? Thank you. -- Shilpa J Bioinformatics Dept., SASTRA University.

1 0