- phenixbb - phenix-online.org

map woes from refine
by Machius, Mischa Christian 08 Mar '13

08 Mar '13

Y'all, Recently, I've run into an issue with the maps from refine. All during the refinement cycles, the autoupdated maps in the associated coot window look great as refinement progresses. However, the final maps look horrible; completely chopped up. It's not a matter of contouring levels; it's essentially unrecognizable, and phenix acknowledges that by stating very low correlation coefficients at the end of the log file. I have downloaded the latest builds in an attempt to address B-value issues that were recently discussed and that I am (still) struggling with. The map issue happens with all of them. Is it possible that there is some crossfire from different versions, or any other reason? Any help appreciated. Cheers, MM

3 4

selection of different residue ranges for a NCS group
by Andres PALENCIA 07 Mar '13

07 Mar '13

Dear All, I am trying to select different regions of the same protein to be included in the same NCS group before refinement. This correspond to the core of a protein. Additionally, I would define other NCS groups for external domains. I have 4 protein/RNA complexes in the asu, P1. Here are the residue ranges for the NCS group corresponding to the core of the protein (invariant): Chain A and resseq: 1:156, 188-227, 413:568, 629:794 this should be the reference group, to match the same region in the other 3 monomers (C, E. G). Other NCS groups will include the RNA, as well as other external domains: Chain B and resseq: 1: 87 This is the RNA molecule, that should match the other 3 monomers (D, F, H) Chain A resseq: 157:187 this is a external domain, to match the other protein monomer (C, E, G) and so on.... for other external domain I have seen that there is the possibility to use user-defined NCS groups, by using: * ncs_groups.params * or modifying the GUI by editing the atom selection For the external domains I think is easy as the syntax is only one line, however for the core of the protein I do not manage to select properly the different residue ranges. Could anybody please help on that? what would be the easiest way to do that? Can I for example creat a ncs_groups.params file and add it to phenix.refine in the GUI? Thanks in advance, Andres

2 1

map scaling: sigma or volume scaled
by Edward A. Berry 06 Mar '13

06 Mar '13

map scaling can be defined: sigma or volume scaled Maybe another dumb question, but what is volume scaling? For 2mFo-DFc and mFo-DFc maps. Thanks, Ed

2 4

quick clashlist or molprobity in batch mode
by jan 06 Mar '13

06 Mar '13

Hi all, I want to analyze a whole bunch of structures with molprobity, and I am looking for a convenient way to do this. phenix.quick_clashlist seems to be a handy tool. However, when I use the following command line: "phenix.quick_clashlist model=mypdb.pdb" ... the output tells me: # run_clashlist # # Run reduce and molprobity # type phenix.run_clashlist: --help for help Parameters used: run_clashlist { model = "4iuj.pdb" build = True summary_file = None output_file = None temp_dir = None remove_temp_dir = None verbose = False raise_sorry = False debug = False dry_run = True } ************************************************ 'str' object has no attribute '_anchor' ************************************************ Any ideas what I need to do differently? thanks, Jan -- Jan Abendroth Emerald BioStructures Seattle / Bainbridge Island WA, USA home: Jan.Abendroth_at_gmail.com work: JAbendroth_at_embios.com http://www.emeraldbiostructures.com

2 1

Using Phenix Refine for Hypothetical Models
by Sam Stampfer 06 Mar '13

06 Mar '13

Dear Phenix Community, Is it possible to use phenix.refine to do simulated annealing and energy minimization of a hypothetical model? I am currently using pyrosetta relax for minimization; however, my hypothetical model is based on a homologous protein whose structure is a crystallographic trimer. I would like to preserve my model's symmetry, which I know is possible in phenix.refine but not in pyrosetta (the symmetry controls are only supported in regular rosetta, as far as I know). Is there a way to have phenix ignore x-ray data and just do energy minimization functions? How would I go about doing that? Would I need a file containing fake/dummy x-ray data? Thanks for your help! -Sam Stampfer

4 4

Fwd: FINAL REMINDER: ALS Call for General User Proposals - DEADLINE March 6 2013
by Banumathi Sankaran 05 Mar '13

05 Mar '13

Dear Users, I am forwarding the reminder mail from ALS to submit your proposal for the CC time and general user time at ALS. Please submit your proposals before time. Thanks Banu FINAL REMINDER: CALL FOR GENERAL USER PROPOSALS AT THE ALS The User Services Office is accepting General User Proposals from scientists who wish to conduct research at the ALS in the next cycle. PROPOSAL SUBMISSION DEADLINE cycle: 2013-2 August-December Deadline: March 6, 2013 NEW PROPOSALS Please follow the instructions for proposal submission at: http://www-als.lbl.gov/index.php/user-information/user-guide/58.html BEAM TIME REQUEST for an ACTIVE PROPOSAL Proposals remain active for two years (up to four cycles), but beam time is only allocated if a beamtime request is made each cycle. After two years the proposal expires and a new proposal must be submitted. If you have an eligible active proposal you will already have received an email from us which includes your login number and password. PUBLICATION REMINDER The proposal form now includes a section on publications from previous ALS work. Please make sure your publications are entered into our database: http://alsusweb.lbl.gov/4DCGI/WEB_GetForm/PublicationSearch.shtml/Initialize If you have questions, please feel free to contact me. Cheers, Clyde Lewis Proposal Coordinator ALS

1 0

Memory and CPU usage
by Jon Agirre 05 Mar '13

05 Mar '13

Dear community, I'm planning on buying a new macbook for my first abroad postdoc venture, and I'm currently trying decide what amount of RAM & CPU to choose. My question is a little bit intricate: is there any rule of thumb to estimate the needed RAM versus the number of input atoms for any particular refinement type? I'm planning on tackling some big (18,000 residues/asu) refinements but I don't have any estimation as to how many GBs of RAM they might require. The maximum amount of RAM I can install is 16GB, but it is a non-standard configuration that might void Apple's warranty. About the CPU: how scalable are typical phenix processes? Would it be sensible to invest in a quad-core machine with HT? In this particular case, and since HT would present 8 logical cores, would I get any speed up from launching phenix tasks configured for 8 processors instead of 4? Thanks in advance! Jon

7 11

Unusually low B factors with phenix 1.8, but not with phenix 1.7 (data set has ice rings)
by Oliv Eidam 04 Mar '13

04 Mar '13

Hi, I would like to continue the B-factor discussion started by Demetres D. Leonidas on Nov 2, 2012 on CCP4BB: http://www.mail-archive.com/[email protected]/msg29012.html I make the same observations like Demetres and Mingzhu Wang: unusually low B-factors when refining with Phenix 1.8-1069 and 1.8.1-1168, but not with Phenix 1.7.1-743. In numbers: the average B-factor is 10 A2 but it should be 20 A2 (based on Wilson B and similar structures). A few atoms (~20) have Bf = 0.00. The following is important: I refined dozens of similar structures without that problem and I observe this phenomenon only for a single data set. That specific data set is different because 1) I excluded regions with ice rings in XDS and 2) I limited the resolution to 2.28 Angstroem, although the crystal diffracted to ~1.8 Angstroem. While the preparation of the data set might be questionable in itself, I would prefer that this would not be the topic of this thread. My questions are: What can I do to refine my structure with Phenix 1.8.X using this truncated data set? What causes the difference in average B-factor values between Phenix 1.7 and 1.8 for this data set? Your advice is much appreciated. Oliv On 11/04/12 19:21, mingzhu wang wrote: > Pavel, > > 1.8.1-1168 also has this problem. > > I found this problem existed with some dataset, but not all dataset. > > 1.8.1 give better map, so I built model with 1.8.1 map and go back to > 1.7.3 before I deposit the structure to PDB. > > When I compared the B factor from 1.8.x and 1.7.3, I found > B(1.7.3)=B(1.8.x)+Boverall(1.8.x). Perhaps this is the problem. > > > > 2012/11/3, Pavel Afonine<pafonine(a)gmail.com>: >> Joao, >> >> I am using the latest version of phenix 1.8.1-1168 >> >> >> 1.8.1-1168 should not have that "problem". >> >> If you suspect there is still a problem, you can send me the data and model >> files off list, explain what exactly the problem is, and I will have a look >> right away. >> >> FYI: there is Phenix mailing list for Phenix-specific questions, where I >> can also explain in great details the behavior with the B-factors you >> observe. >> >> All the best, >> Pavel >> -- Oliv Eidam, Ph.D. Postdoctoral fellow University of California, San Francisco Dept. of Pharmaceutical Chemistry 1700 4th Street, Byers Hall North, Room 501 San Francisco, CA 94158 - Box 2550 Phone: 415-514-4253 Fax : 415-514-4260 Email:[email protected]

3 5

fwd: Reminder: CCP4 summer school at APS, USA
by Nathaniel Echols 04 Mar '13

04 Mar '13

---------- Forwarded message ---------- From: "Ruslan Sanishvili (Nukri)" <nukri(a)aps.anl.gov> To: phenixbb(a)phenix-online.org Date: Mon, 4 Mar 2013 16:14:43 -0600 (CST) Subject: Reminder: CCP4 summer school at APS, USA Dear Colleagues, This is a reminder that the deadline for applications for the 6th annual CCP4 Summer School “From data collection to structure refinement and beyond” is April 5, 2013. The school will take place from June 18 through June 26, 2013 at the Advanced Photon Source (APS) near Chicago. There is no registration fee for the school. The students will be responsible for their own travel and lodging expenses. These and other details (the program, the list of speakers, the application process, accommodations, site access, contacts etc) can be found at the workshop website at http://www.ccp4.ac.uk/schools/APS-2013/index.php The school will include data collection, processing, structure solution, model building, refinement, validation, automation of many steps etc. Participants are encouraged to bring their own crystals, raw data or processed data for hands-on problem solving under the guidance of software developers and other experts. We hope to see you in June Garib, Ronan and Nukri Ruslan Sanishvili (Nukri) Macromolecular Crystallographer GM/CA@APS X-ray Science Division, ANL 9700 S. Cass Ave. Argonne, IL 60439 Tel: (630)252-0665 Fax: (630)252-0667 rsanishvili(a)anl.gov

1 0

Bay Area and Beyond Phenix User Meeting (March 14th)
by Paul Adams 04 Mar '13

04 Mar '13

Dear Colleagues, On Thursday March 14th, a Bay Area and Beyond Phenix User Meeting will be held at Lawrence Berkeley Lab. During the workshop, the Phenix developers will give lectures on the latest developments and Phenix usage. Half of the user meeting will be in the form of informal hands on sessions, allowing participants to work with the developers. Participants are encouraged to bring their own (reduced) data to be used during hands on sessions, but tutorial data will also be available. We also recommend bringing a relatively recent laptop computer (any OS). Due to the highly interactive format of the meeting there will be space for approximately 20 participants. If I have missed any of your colleagues on this invitation please feel free to pass on the message. To apply, researchers should send a short summary email describing their level of expertise and motivation for attending the workshop to NEchols(a)lbl.gov and PDAdams(a)lbl.gov by Friday March 8th. Cheers, Paul -- Paul Adams Deputy Division Director, Physical Biosciences Division, Lawrence Berkeley Lab Division Deputy for Biosciences, Advanced Light Source, Lawrence Berkeley Lab Adjunct Professor, Department of Bioengineering, U.C. Berkeley Vice President for Technology, the Joint BioEnergy Institute Laboratory Research Manager, ENIGMA Science Focus Area Building 64, Room 248 Tel: 1-510-486-4225, Fax: 1-510-486-5909 http://cci.lbl.gov/paul Lawrence Berkeley Laboratory 1 Cyclotron Road BLDG 64R0121 Berkeley, CA 94720, USA. Executive Assistant: Louise Benvenue [ LBenvenue(a)lbl.gov ][ 1-510-495-2506 ] --

1 0

phenix.refine multiple twin laws?
by Kip Guja 03 Mar '13

03 Mar '13

Hi All, Is the newest version of Phenix capable of refinement with multiple twin laws (eg, tetartohedral twinning)? If so, any advice on input parameters, etc. would be greatly appreciated. Best wishes, Kip

2 1

Re: [phenixbb] Magnesium or calcium
by Daniel Slade 02 Mar '13

02 Mar '13

So the resolution is 1.6 and I remember the B factor for mg at 16. I haven't checked the coordinating residues. It coordinates acidic residues, a backbone carbonyl and one water. I'm now leaning towards calcium as the other calciums have bonds distances similar to this one, but not nearly as high an affinity. Thanks

2 1

Magnesium or calcium
by Daniel Slade 02 Mar '13

02 Mar '13

I have several structures that has multiple known calcium binding sites, and I found a new site that seems to accept calcium or magnesium. I say this because if a magnesium is put in the structure, the bond lengths are ~ 2.1. But if you put a calcium in it, they go to ~2.6. All of these refinements were done without ReadySet, which I recently used with metal restraints and it seemed to change the bond length of calcium to ~2.4. I'm a bit confused as to what to do here as either metal seems to be right. Worth noting, this is a very strong metal binding site that can't be removed with 10 mM EDTA. Any help here on how to figure this out which ion this is would be great. Thanks in advance Dan

2 1

Re: [phenixbb] Magnesium or calcium
by Ed. Pozharski 02 Mar '13

02 Mar '13

Magnesium has fewer electrons. So if you place it and it's actually calcium you should see extra density and the B-factor will be unreasonably lower than surrounding atoms. Distances are less reliable at intermediate resolutions as they are largely controlled by restraints. Maybe. Assuming that you have 100% occupancy. -------- Original message -------- From: Daniel Slade <DSlade(a)scripps.edu> Date: To: phenixbb(a)phenix-online.org Subject: [phenixbb] Magnesium or calcium I have several structures that has multiple known calcium binding sites, and I found a new site that seems to accept calcium or magnesium. I say this because if a magnesium is put in the structure, the bond lengths are ~ 2.1. But if you put a calcium in it, they go to ~2.6. All of these refinements were done without ReadySet, which I recently used with metal restraints and it seemed to change the bond length of calcium to ~2.4. I'm a bit confused as to what to do here as either metal seems to be right. Worth noting, this is a very strong metal binding site that can't be removed with 10 mM EDTA. Any help here on how to figure this out which ion this is would be great. Thanks in advance Dan _______________________________________________ phenixbb mailing list phenixbb(a)phenix-online.org http://phenix-online.org/mailman/listinfo/phenixbb

1 0

question about reflection file tools behavior
by Michael Thompson 01 Mar '13

01 Mar '13

Hello All, I have a question about the behavior of the reflection file tools within the PHENIX GUI: I am trying to simply copy a set of FreeRFlags from a higher resolution dataset to a lower resolution dataset. The low-res dataset contains 35264 reflections in the original input mtz. The high-res set contains 43731 reflections. When I use the GUI to write a new mtz using the low-res amplitudes and the high-res FreeRFlags, the result is an mtz with 46323 reflections. I am confused about why the output has so many reflections, even more than the high res dataset? Also, I noticed that while both input mtz files have a low-res cutoff of 20A, the output mtz has a low-res cutoff around 75A. Can anyone please explain what is going on here? I would like my output file to have the same reflections as the input, and no more. I have noticed that the CCP4 program FreeRFlag seems to behave in a similar way, so my guess is that what is happening is a common procedure, but I'm at a loss for an explanation of why this is the case. Also, if there's a better way to implement what I am trying to do I'd be grateful to hear it (perhaps generate entirely new flags for both datasets using the same random seed?). Thanks for the help!! Mike -- Michael C. Thompson Graduate Student Biochemistry & Molecular Biology Division Department of Chemistry & Biochemistry University of California, Los Angeles miket(a)chem.ucla.edu

2 1

mr_rosetta fails after autobuild stage
by Huw Jenkins 01 Mar '13

01 Mar '13

Hi, I've been having some problems with mr_rosetta. All the regression tests run fine but when I run it on some real data it gets as far as the autobuilding stage and then fails with an error message. All I can find in the log files GROUP_OF_AUTOBUILD_1/RUN_FILE_1.log GROUP_OF_AUTOBUILD_1/RUN_FILE_2.log is ******************* ERROR ENDING *************** Failed to carry out autobuild ******************* ERROR ENDING *************** The autobuild jobs themselves seem to finish fine i.e the end of the log files in GROUP_OF_AUTOBUILD_1/RUN_1/AutoBuild_run_1_ and GROUP_OF_AUTOBUILD_1/RUN_2/AutoBuild_run_1_ has the tables of Rfactors per cycle and the list of citations with no indiction of any error. Is there anywhere else I should be looking for a more detailed error message? Thanks, Huw

2 4

real_space_correlation
by Jianghai Zhu 28 Feb '13

28 Feb '13

Hi All, I was trying to use "phenix.real_space_correlation *.pdb resolve_composite_map.mtz" to get some real space cc for my ligand against the SA composite omit map. But I am not sure that the map used here was the SA composite omit map. Will phenix use the omit map I provided or will it just simply calculate a sigmaA weighted 2Fo-Fc map based on the pdb? Thanks. --Jianghai

2 4

mr_rosetta: two proteins, one chain placed, build second chain
by Tobias Beck 28 Feb '13

28 Feb '13

Dear phenixbb, I would like to use mr_rosetta to build a second protein chain. For the first protein, I have a search model with known structure. The search model can be placed with Phaser (TFZ 10). For the second protein, the structure of a homologue has been determined. The second protein accounts for about 2/3 of the residues in the ASU. I would like to fix the first protein chain and build the second protein chain with mr_rosetta. I have obtained fragment files for both proteins from robetta. What input parameters are required to tell mr_rosetta to build the second chain only? I have tried the fixed_ensemble_ID_list = ensemble1 option. Here is an excerpt of my parameter file with the input files: run a) search_models = "path/chainA.pdb" (from Phaser) search_models = "path/chainB.pdb" (homologue structure) fixed_ensembles { fixed_ensembleID_list = ensemble1 fixed_euler_list = 0.0 0.0 0.0 fixed_frac_list = 0.0 0.0 0.0 fixed_frac_list_is_fractional = True } alignment_files = "path/chainA.dat" alignment_files = "path/chainB.dat" with chainA.dat having twice the same sequence and chainB.dat with the target and template alignment. I have set fragment files with chain_list for both chains. I get this error when running phenix.mr_rosetta (running from command line with .eff file) "Sorry, the file /path/GROUP_OF_PLACE_MODEL_1/RUN_2/chainB_al_0.pdb representing group 0 has no atoms? This could indicate a problem with sequence alignment in rebuilding. It could be that there are near-duplicate sequences in your sequence file ...if this is the case then you may need to artificially rename your duplicated chains and renumber residues to make a single chain containing a duplication" This error also occurs run b) Without any alignment files run c) Without any alignment files and with just chainA.pdb as search model. An alignment of the sequences from chainA and chainB gives low identity (no duplicate regions), so I suspect the problem is located somewhere else? I have seen a post by Tom Terwilliger, explaining how to deal with two identical chains with different conformations where no alignment files are required. However, I would like to use the information from the homologue structure, that is why I specified the alignment file. By inputting chainB.pdb as search model, mr_rosetta also tries to use this model as a search model for molecular replacement, but I think this is not necessary (and gives a warning that the space group of the search model and the data file do not match). I just want to provide the structural information from structure chainB.pdb to mr_rosetta for building. ----- So in summary this is what I would like to do: chain A: keep fixed - use chainA.pdb as fixed model chain B: build - use information from homologue structure (chainB.pdb) and fragment files to build chain What is the best strategy to do this? What input parameters are required so that I do not get the error described above? ----- Thanks a lot in advance, Tobias.

3 2

phaser output and packing
by Ursula Schulze-Gahmen 26 Feb '13

26 Feb '13

Could somebody explain to me how phaser decides if there is a solution. When I look at the log file there are a few solutions in the translation function that have fewer than the allowed clashes, but they are still not recognized as potential solutions. Why is that. Is there another parameter that I should be looking at? Ursula -- Ursula Schulze-Gahmen, Ph.D. Assistant Researcher UC Berkeley, QB3 356 Stanley Hall #3220 Berkeley, CA 94720-3220

3 2

WCPCW 21
by Williams, John Charles 22 Feb '13

22 Feb '13

All, We will be finalizing the program for the workshop on Monday, Feb 25th. The Monterey Plaza and Spa has confirmed that we can accommodate a few more people. So, if you are interested in attending and would like to be considered for an oral presentation or would like to present a poster, please register and submit an abstract no later than Monday, 12 noon (PST). The meeting will be March 17th-20th with Wayne Hendrickson as the Keystone speaker! Best, John wcpcw.org --------------------------------------------------------------------- *SECURITY/CONFIDENTIALITY WARNING: This message and any attachments are intended solely for the individual or entity to which they are addressed. This communication may contain information that is privileged, confidential, or exempt from disclosure under applicable law (e.g., personal health information, research data, financial information). Because this e-mail has been sent without encryption, individuals other than the intended recipient may be able to view the information, forward it to others or tamper with the information without the knowledge or consent of the sender. If you are not the intended recipient, or the employee or person responsible for delivering the message to the intended recipient, any dissemination, distribution or copying of the communication is strictly prohibited. If you received the communication in error, please notify the sender immediately by replying to this message and deleting the message and any accompanying files from your system. If, due to the security risks, you do not wish to receive further communications via e-mail, please reply to this message and inform the sender that you do not wish to receive further e-mail from the sender. (fpc5p) ---------------------------------------------------------------------

1 0

Maximum Entropy Maps
by Morten Groftehauge 22 Feb '13

22 Feb '13

Hi guys, I see that phenix.maximum_entropy_map is now a command in Phenix. Some quick questions: Where is this likely to be must useful and does it take ridiculously long to run? From the Nishibori 2008 paper in Acta D it seems like this would mainly be useful for very high resolution structures that you would normally call complete - and that it would take a very long time to compute. I love trying out new methods but I need a bit more to go on with this one. Cheers, Morten -- Morten K Grøftehauge, PhD Pohl Group Durham University

5 7

Building sugars
by Folmer Fredslund 21 Feb '13

21 Feb '13

Dear all, What's the "correct" way to build and refine sugar polymers? I am currently building several structures with different kinds of sugar polymers bound to them. Searching for similar "ligands" in the PDB, I end up with e.g. trisaccharides that are named as one molecule, even though they are indeed made up of three individual sugars with bonds between them. Thank you for any pointers. Best regards, Folmer -- Folmer Fredslund

1 0

phenix.table_one citation
by Alexander Batyuk 20 Feb '13

20 Feb '13

Dear developers, How to cite phenix.table_one? Best wishes, Alex -- Alex Batyuk The Plueckthun Lab www.bioc.uzh.ch/plueckthun

2 1

cut out density vs FFT
by Yunji Wu 19 Feb '13

19 Feb '13

Hi all-- I'm trying to make a 2Fo-Fc map with density wrapped around my asymmetric unit to make a figure showing density for my sub-8Å structure. I am not building into this map or doing MR. My model was refined in Phenix.refine (R's in the high 30's), and so I have the output model .pdb and the output map coefficients .mtz. When I input these into cut_out_density, the only "map labels" input options it gives me are 2FOFCWT, PH2FOFCWT, which seem fine. Using either box, sphere, or model for the cutout type, I get a VERY nice looking map that looks extremely solvent flattened and tight around my model, albeit still to 8Å. Suspicious of anything that looks good, I went to try the same thing in FFT (CCP4). There I generated a simple map (using the same .mtz) to cover all atoms in a pdb file- the same model I put into cut_out_density. For the input I selected 2FOFCWT and PH2FOFCWT, leaving Sigma and Weight unassigned. The map that comes out of this looks terrible, with a lot of noise and little density corresponding to the model aside from what one might expect from a very poor resolution experimental map. I've also tried using FFT in Phenix to do the same thing, and it gives like better maps than CCP4 FFT but worse than cut_out_density. It almost appears as if there is some additional solvent flattening going on, which I know happens at the beginning of refinement, but that should be intrinsic to the map coefficients file which is common to everything I've tried. My questions are (1) what should I use for what I am trying to do (just display 2Fo-Fc density around asymmetric unit for a figure) and (2) what is the difference between what these three programs (cut_out, FFT-CCP4, FFT-Phenix) are calculating? The inputs are exactly the same -- mtz (2FOFCWT and PH2FOFCWT) and pdb output from refinement. Thanks for the help, Yunji

3 2

Problems placing second component in Phaser
by Morten Groftehauge 19 Feb '13

19 Feb '13

Hi guys, So I have a solution that covers some of target and I ran it through phenix.mr_rosetta and I am actually kinda happy with the output from there. However, while I can place both (NCS=2) with the original model I only seem to be able to place one of the rosetta models, even with 'packing function' set to 'all' or 'allow'. Original model - Refinement after placing component 1: LLG 175 Refinement after placing component 2: LLG 620 Rosetta model (packing 'all') - Refinement after placing component 1: LLG 295 Refinement after placing component 2: LLG 390 Rosetta model (packing standard) - Refinement after placing component 1: LLG 295 Refinement after placing component 2: LLG 350 More notably there is little to no increase in TFZ with the placement of the second component... There are more residues in the Rosetta model and I would expect more clashes but this looks weird. Cheers, Morten -- Morten K Grøftehauge, PhD Pohl Group Durham University

3 3