- phenixbb - phenix-online.org

AQuaRef - issues with torch-cluster installation
by Martin Malý 16 Oct '24

16 Oct '24

Dear colleagues, I'm trying to install AQuaRef. However, I'm struggling with running `bash config/update_phenix.sh aimnet2` which is part of `bash build_into_phenix.sh aimnet2`. This script wants to install the python package torch-cluster. However, I'm getting the error: /opt/phenix/phenix-dev-5480/conda_base/lib/python3.9/site-packages/torch/include/torch/csrc/python_headers.h:12:10: fatal error: Python.h: No such file or directory 12 | #include <Python.h> | ^~~~~~~~~~ compilation terminated. error: command '/usr/bin/gcc' failed with exit code 1 [end of output] I'm getting the same error when I tried to install only this package using the command `qrefine.python -m pip install torch-cluster`. I have a laptop with Ubuntu 22.04. I have all the usual development packages installed (build-essential, linux-headers, gcc, g++, ...). According Python, I have libpython3-dev, libpython3.10-dev, python3-dev, python3.10-dev. I can see that qrefine.python has a version of 3.9, so I also installed libpython3.9-dev, python3.9-dev and python3.9-distutils from ppa:deadsnakes/ppa but it did not help. I am also attaching a full log from the torch-cluster installation. It seems that the package torch has been installed successfully in prior. Do you have please an idea what is wrong and how to fix it? Thank you. Martin

2 4

help for the 4.3 A dataset
by princeadam＠163.com 15 Oct '24

15 Oct '24

Dear colleagues, I am currently engaged in an endeavor to unravel the crystal structure through molecular replacement techniques. Regrettably, the resolution attained stands at a rather low mark, approximately 4.3 Angstroms, which poses significant challenges. Despite exhaustive efforts directed towards optimizing the crystal conditions, there has been no discernible improvement in the resolution. The structure in question is confined to a specific domain, and we have at our disposal two distinct conformational models sourced from the PDB database: one derived from X-ray data and the other from Cryo-EM. Our prime objective is to meticulously discern which of these two conformations aligns most accurately with our experimental context. To this end, we have employed both structures (X-ray and Cryo-EM) as templates for molecular replacement, yet the refinement process has hit a roadblock, hindering our progress. We are seeking a collaborator with experience in dealing with low-resolution data to assist in solving the structure. We are open to sharing first or corresponding authorships in the near future for this paper. Warm regards Hui

2 1

NIH would like your input
by Paul Adams 15 Oct '24

15 Oct '24

Dear Colleagues, For those of you in the US, you might want to know that NIH issued a Request for Information about their National Resources and Biomedical Technology Optimization and Dissemination programs: https://www.research.net/r/NRRs_and_Biomedical_Technology_Optimization_and_… Unfortunately, it closes tomorrow, but it is a short form. They are interested in experiences with existing resources, but are most interested in hearing ideas for expanding the programs to new technologies. Currently funded resources can be found here: https://www.nigms.nih.gov/Research/mechanisms/Pages/NIGMS-National-and-Regi… Information about the Biomedical Technology Optimization and Dissemination program can be found here: https://www.nigms.nih.gov/grants/Pages/btod.aspx NIH would appreciate your thoughts about additional technologies that would benefit the community by support through these programs. Cheers, Paul -- Paul Adams (he/him/his) Associate Laboratory Director for Biosciences, LBL (https://biosciences.lbl.gov) Principal Investigator, Computational Crystallography Initiative, LBL (http://cci.lbl.gov) Vice President for Technology, the Joint BioEnergy Institute (http://www.jbei.org) Principal Investigator, ALS-ENABLE, Advanced Light Source (http://als-enable.lbl.gov) Laboratory Research Manager, ENIGMA Science Focus Area (http://enigma.lbl.gov) Adjunct Professor, Department of Bioengineering, UC Berkeley (http://bioeng.berkeley.edu) Member of the Graduate Group in Comparative Biochemistry, UC Berkeley (http://compbiochem.berkeley.edu) Building 91, Room 410 Building 978, Room 4126 Tel: 1-510-486-4225 http://cci.lbl.gov/paul ORCID: 0000-0001-9333-8219 Lawrence Berkeley Laboratory 1 Cyclotron Road BLDG 91R0183 Berkeley, CA 94720, USA. Executive Assistant: Michael Espinosa [ MEEspinosa(a)lbl.gov ][ 1-510-333-6788 ] Phenix Consortium: Ashley Dawn [ AshleyDawn(a)lbl.gov ][ 1-510-486-5455 ] --

1 0

NIH would like your input
by Paul Adams 15 Oct '24

15 Oct '24

Dear Colleagues, For those of you in the US, you might want to know that NIH issued a Request for Information about their National Resources and Biomedical Technology Optimization and Dissemination programs: https://www.research.net/r/NRRs_and_Biomedical_Technology_Optimization_and_… Unfortunately, it closes tomorrow, but it is a short form. They are interested in experiences with existing resources, but are most interested in hearing ideas for expanding the programs to new technologies. Currently funded resources can be found here: https://www.nigms.nih.gov/Research/mechanisms/Pages/NIGMS-National-and-Regi… National Institute of General Medical Sciences nigms.nih.gov Information about the Biomedical Technology Optimization and Dissemination program can be found here: https://www.nigms.nih.gov/grants/Pages/btod.aspx NIH would appreciate your thoughts about additional technologies that would benefit the community by support through these programs. Cheers, Paul -- Paul Adams (he/him/his) Associate Laboratory Director for Biosciences, LBL (https://biosciences.lbl.gov) Principal Investigator, Computational Crystallography Initiative, LBL (http://cci.lbl.gov) Vice President for Technology, the Joint BioEnergy Institute (http://www.jbei.org) Principal Investigator, ALS-ENABLE, Advanced Light Source (http://als-enable.lbl.gov) Laboratory Research Manager, ENIGMA Science Focus Area (http://enigma.lbl.gov) Adjunct Professor, Department of Bioengineering, UC Berkeley (http://bioeng.berkeley.edu) Member of the Graduate Group in Comparative Biochemistry, UC Berkeley (http://compbiochem.berkeley.edu) Building 91, Room 410 Building 978, Room 4126 Tel: 1-510-486-4225 http://cci.lbl.gov/paul ORCID: 0000-0001-9333-8219 Lawrence Berkeley Laboratory 1 Cyclotron Road BLDG 91R0183 Berkeley, CA 94720, USA. Executive Assistant: Michael Espinosa [ MEEspinosa(a)lbl.gov ][ 1-510-333-6788 ] Phenix Consortium: Ashley Dawn [ AshleyDawn(a)lbl.gov ][ 1-510-486-5455 ] --

1 0

Building Rosetta for Phenix
by Martin Malý 14 Oct '24

14 Oct '24

Dear colleagues, I am dealing with a low resolution data set (~4 A, protein crystal structure) so I would like to explore phenix.rosetta_refine. I followed the instructions at https://phenix-online.org/documentation/reference/rosetta_install.html . I applied the patch in ${PHENIX_ROSETTA_PATH}/main/source/tools/build/site.settings . After several minutes of compilation (command rosetta.build_phenix_interface), I got an error. I re-run the command to get a log file which I'm attaching. Please don't you have any idea what's wrong? I use Ubuntu 22.04. Thank you. Best regards, Martin

3 3

Temperature shift alters unit cell of insulin: an issue
by Esra A 09 Oct '24

09 Oct '24

Dear Phenix community, I performed a temperature shift data collection for di-hexamer insulin. I collected single crystal data at 100K, 200K, and 300K, respectively After processing the data, I noticed that the unit cell of the 100K dataset is distinct from 200K and 300K. 100K dataset = 78.2182 78.2182 79.3565 90 90 120 (assume that this is di-hexamer) 200K dataset = 78.3157 78.3157 39.6367 90 90 120 (it is hexamer) 300K dataset = 80.7109 80.7109 39.6399 90 90 120 (it is hexamer) Even if I completed the structure determination of 200K and 300K successfully, I couldn't complete the molecular replacement of the 100K dataset due to the "*The composition entered will not fit in the unit cell volume*" error message. I tried every "copies" and "mass" parameter in *automr *script. However, i got the error message every run. I can share with you the mtz and reference pdb file, if needed. Could you help me, please? Where am I making a mistake? Thank you very much. Sincerely, Esra Ayan

4 5

real space refine ADPs
by Leonid Sazanov 08 Oct '24

08 Oct '24

Hi, according to online description phenix.real_space_refine should refine only group ADPs per residue. However, in practice after the run with ADP included in strategy the recent versions refine individual B-factors per atom. They seem to look OK, but I wonder if this is a beta feature or fully implemented by now? Many thanks Leonid -- Prof. Leonid Sazanov FRS Institute of Science and Technology Austria (ISTA) Am Campus 1 A-3400 Klosterneuburg Austria Phone: +43 2243 9000 3026 E-mail: sazanov(a)ist.ac.at

3 2

new BUILT of XDS
by kay.diederichs＠uni-konstanz.de 04 Oct '24

04 Oct '24

Dear XDS users in the Phenix community, as reported by users, and confirmed in a larger-scale comparison by GlobalPhasing, the BUILT=20240723 of XDS Version 20240630 leads to problems for some datasets when compared with (the expired) XDS Version 20230630. (it may be better than the 20230630 version for other datasets!). An improved BUILT=20241002 of the 20240630 version is available (all platforms) from https://xds.mr.mpg.de/html_doc/downloading.html for non-commercial (academic) users (commercial users have their own arrangements). In my/our testing, it is at least as good as version 20230630, and often better. Please install and use this instead of BUILT=20240723. If you still find deficiencies, please contact me or Wolfgang Kabsch - but be prepared to share the raw data (confidentially, of course). To enable comparisons, the old XDS version 20230630 was re-released for Linux by Wolfgang Kabsch with expiration date 2025-Mar-31; the link to its XDS_old.tar.gz is at https://xds.mr.mpg.de/ . Best wishes, Kay

3 2

Registration now open for CCP4 Study Weekend 2025 - Using software, AI and other methods to advance macromolecular models
by Karen McIntyre - STFC UKRI 03 Oct '24

03 Oct '24

Dear all, Registration is now open for the 2025 CCP4 study weekend entitled " Using software, AI and other methods to advance macromolecular models". This year the CCP4 Study Weekend will be held as a hybrid event from the 7th to 9th January 2025, enabling people to choose whether to attend in-person or virtually. The in-person event will be held at the East Midlands Conference Centre, Nottingham, UK. We would like to invite you to another iteration of the ever-popular CCP4 Study Weekend - to start the year 2025 with a lot of fresh ideas, new insights and (hopefully) new friends and contacts to boost. As always it will be an eclectic mix of bleeding-edge science, in-depth presentations, discussion panel, poster-sessions, hands-on tutorials and plenty of opportunities for social interactions. For full details, programme, logistics and registration, please visit the Study Weekend website<https://studyweekend.ccp4.ac.uk/>. There is no better way to start 2025 than to put participation in that exciting meeting at the top of your "New Year's Resolutions" list! Key Dates: * Early bird registration final date: 11 November 2024; * Standard Student bursaries are open to all students registering during early bird period i.e. now until 11 November which cover the cost of registration plus one night accommodation in halls (1 night in hotel will be subsidised for those students choosing to stay in the hotel); * Travel bursaries are available for students and young postdocs studying at overseas labs to help cover cost of travel to UK. The deadline for applications is 31 October 2024; * Registration for in-person delegates closes 4 December 2023 (or earlier if in-person places sell out although you will be able to join the in-person waitlist). * Registration for virtual delegates will close 9 January 2025. Key Timings Diamond MX user meeting will start at 11:00 on 7TH January 2025 CCP4 study weekend 2025 will start at ~17:30 on 7th January 2025 the event will finish at ~16:30 on 9th January 2025 We hope to see you there! Scientific Organisers: Elke De Zitter (Institut de Biologie Structurale, FRANCE) Deborah Harrus (EMBL-EBI, UK) Dorothee Liebschner (Lawrence Berkeley National Laboratory, USA) Administrative Organisers: Karen McIntyre (CCP4, UK)

1 0

ResolveCryoEM inputs
by Josh Cofsky 17 Sep '24

17 Sep '24

Hi, I'm attempting to run ResolveCryoEM, and I have a few questions about setting the job up correctly. My half-maps were reconstructed with C4 symmetry to ~2 Å (CryoSPARC), and my model was refined into the density with the "Ncs constraints" box checked (Phenix RSR). My sample is a protein-DNA complex (4 copies of a protein protomer, 4 copies of a ssDNA protomer). I'm running Phenix 1.21.1-5286-000 (the latest installation provided by SBGrid). In the first version of the run, I only supplied the two half-map files and a seq file. Question 1: My current seq.dat file looks like this (as mentioned previously in the forum, I was getting errors when I tried the version where my seq.dat file has only 1 protomer and I set the number of copies to 4 in the GUI field "Copies of sequence file in map"): >protein_protomer1 MGSRSG... >protein_protomer2 MGSRSG... >protein_protomer3 MGSRSG... >protein_protomer4 MGSRSG... >DNA_protomer1 AAAAAA >DNA_protomer2 AAAAAA >DNA_protomer3 AAAAAA >DNA_protomer4 AAAAAA Currently, Phenix is reading in the "AAAAAA" as a protein chain ("Chain types considered: PROTEIN"). How do I tell Phenix that those are deoxyadenosines, not alanines? Question 2: Is Phenix automatically detecting the C4 symmetry and density-modifying with that constraint? The resulting map looks symmetric in the noise, but I don't see anything in the log suggesting that it knows the symmetry. Here are some settings from the log that seem relevant (all left to default values): refine_symmetry_in_denmod = True use_symmetry_in_denmod = False symmetry_in_denmod_last_cycle_only = True fraction_ncs = None I attempted to supply a symmetry file (the output of a MapSymmetry job on the full map = "symmetry_from_map.ncs_spec"), but the file fails to load into the input ("Phenix error: Phenix did not recognize the file type for ****/symmetry_from_map.ncs_spec"). Question 3: A part of my protein is disordered and does not appear in the cryo-EM density (and was thus left unmodeled). Is it correct to leave that piece out of the seq.dat file? Question 4: If I include a model in the input, Phenix demands a symmetry file, but I couldn't get my symmetry file to load as explained above. What is the symmetry file supposed to look like? Thank you! -Josh

2 1