Configuring ChimeraX AlphaFold “Fetch” and “Search” features
Hi ChimeraX Team In our commercial lab we are unable to access public domain AlphaFold databases, and instead have our own copy on-premises. Is there a mechanism to reconfigure the ChimeraX AlphaFold “Fetch” and “Search” features to query our in-house copy? ChimeraX error reported is: Fetching url https://alphafold.ebi.ac.uk/files/AF-Q5STB3-F1-model_v4.cif failed: HTTP Error 403: Forbidden Many Thanks for advising me Best Regards Jim [cid:image001.png@01DB1A36.151649E0] Jim Thorpe Scientific Investigator Protein, Cellular & Structural Sciences Target Discovery GSK Medicines Research Centre, Gunnels Wood Road, Stevenage, SG1 2NY, UK james.h.thorpe@gsk.com [cid:image002.png@01DB1A36.151649E0]<https://www.gsk.com/> [cid:image003.png@01DB1A36.151649E0]<https://www.facebook.com/GSK> [cid:image004.png@01DB1A36.151649E0]<https://twitter.com/GSK> [cid:image005.png@01DB1A36.151649E0]<https://www.linkedin.com/company/glaxosmithkline> GSK monitors email communications sent to and from GSK in order to protect GSK, our employees, customers, suppliers and business partners, from cyber threats and loss of GSK Information. GSK monitoring is conducted with appropriate confidentiality controls and in accordance with local laws and after appropriate consultation.
Hi Jim, The AlphaFold database fetch in ChimeraX uses an https query to the EBI database as you've seen. It does not have any provision to get the file from local disk. Most ChimeraX users do not have a copy of the 200 million AlphaFold database structures locally, but I understand why GSK wants that. If your local copy of AlphaFold database is accessible via https within your company I could tell you how to make ChimeraX use it. If your copy is just a bunch of files on a file system, I could also tell you how to modify a couple lines of ChimeraX Pythoin code to use those files. Or you could just use the ChimeraX open command to open the AlphaFold database file (e.g. "open /path/to/AFDB/files/AF-P26039-F1-model_v4.cif"). The AlphaFold database Search capability runs using a web service at UC San Francisco that we wrote to do a fast k-mer search of the 200 million sequences. I'm sure you don't have that at GSK. It requires a custom built sequence k-mer database. It also is only designed to run as a REST web service and would probably be a good bit of trouble for you to setup at GSK. Tom
Hi Tom If you could share the https configuration details this would be perfect 😊 Many Thanks Jim From: Tom Goddard <goddard@sonic.net> Sent: 09 October 2024 17:12 To: Jim Thorpe <james.h.thorpe@gsk.com> Cc: chimerax-users@cgl.ucsf.edu Subject: Re: [chimerax-users] Configuring ChimeraX AlphaFold “Fetch” and “Search” features Hi Jim, The AlphaFold database fetch in ChimeraX uses an https query to the EBI database as you've seen. It does not have any provision to get the file from local disk. Most ChimeraX users do not have a copy of the 200 million AlphaFold database Hi Jim, The AlphaFold database fetch in ChimeraX uses an https query to the EBI database as you've seen. It does not have any provision to get the file from local disk. Most ChimeraX users do not have a copy of the 200 million AlphaFold database structures locally, but I understand why GSK wants that. If your local copy of AlphaFold database is accessible via https within your company I could tell you how to make ChimeraX use it. If your copy is just a bunch of files on a file system, I could also tell you how to modify a couple lines of ChimeraX Pythoin code to use those files. Or you could just use the ChimeraX open command to open the AlphaFold database file (e.g. "open /path/to/AFDB/files/AF-P26039-F1-model_v4.cif"). The AlphaFold database Search capability runs using a web service at UC San Francisco that we wrote to do a fast k-mer search of the 200 million sequences. I'm sure you don't have that at GSK. It requires a custom built sequence k-mer database. It also is only designed to run as a REST web service and would probably be a good bit of trouble for you to setup at GSK. Tom On Oct 9, 2024, at 3:10 AM, Jim Thorpe via ChimeraX-users <chimerax-users@cgl.ucsf.edu<mailto:chimerax-users@cgl.ucsf.edu>> wrote: Hi ChimeraX Team In our commercial lab we are unable to access public domain AlphaFold databases, and instead have our own copy on-premises. Is there a mechanism to reconfigure the ChimeraX AlphaFold “Fetch” and “Search” features to query our in-house copy? ChimeraX error reported is: Fetching url https://alphafold.ebi.ac.uk/files/AF-Q5STB3-F1-model_v4.cif<https://urldefense.com/v3/__https:/alphafold.ebi.ac.uk/files/AF-Q5STB3-F1-model_v4.cif__;!!AoaiBx6H!xgCTEqyyRjjbm29BS6hWMOzVlNXvKpm3fkurJ6jjGa8a4Rq6sqPzW6ElqY2umAEImZIXhA7tyEGTFv70xg$> failed: HTTP Error 403: Forbidden Many Thanks for advising me Best Regards Jim <image001.png> Jim Thorpe Scientific Investigator Protein, Cellular & Structural Sciences Target Discovery GSK Medicines Research Centre, Gunnels Wood Road, Stevenage, SG1 2NY, UK j<mailto:james.h.thorpe@gsk.com>ames.h.thorpe@gsk.com<mailto:james.h.thorpe@gsk.com> <image002.png><https://www.gsk.com/> <image003.png><https://urldefense.com/v3/__https:/www.facebook.com/GSK__;!!AoaiBx6H!xgCTEqyyRjjbm29BS6hWMOzVlNXvKpm3fkurJ6jjGa8a4Rq6sqPzW6ElqY2umAEImZIXhA7tyEFi-01njA$> <image004.png><https://urldefense.com/v3/__https:/twitter.com/GSK__;!!AoaiBx6H!xgCTEqyyRjjbm29BS6hWMOzVlNXvKpm3fkurJ6jjGa8a4Rq6sqPzW6ElqY2umAEImZIXhA7tyEHXtn_a9g$> <image005.png><https://urldefense.com/v3/__https:/www.linkedin.com/company/glaxosmithkline__;!!AoaiBx6H!xgCTEqyyRjjbm29BS6hWMOzVlNXvKpm3fkurJ6jjGa8a4Rq6sqPzW6ElqY2umAEImZIXhA7tyEH9lpTLsQ$> GSK monitors email communications sent to and from GSK in order to protect GSK, our employees, customers, suppliers and business partners, from cyber threats and loss of GSK Information. GSK monitoring is conducted with appropriate confidentiality controls and in accordance with local laws and after appropriate consultation. _______________________________________________ ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu<mailto:chimerax-users@cgl.ucsf.edu> To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu<mailto:chimerax-users-leave@cgl.ucsf.edu> Archives: https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/<https://urldefense.com/v3/__https:/mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/__;!!AoaiBx6H!xgCTEqyyRjjbm29BS6hWMOzVlNXvKpm3fkurJ6jjGa8a4Rq6sqPzW6ElqY2umAEImZIXhA7tyEEhV2xKHA$> GSK monitors email communications sent to and from GSK in order to protect GSK, our employees, customers, suppliers and business partners, from cyber threats and loss of GSK Information. GSK monitoring is conducted with appropriate confidentiality controls and in accordance with local laws and after appropriate consultation.
Hi Jim, There is a configuration file in the user settings directory for ChimeraX named alphafold_database-1 that contains the URL pattern for fetching AlphaFold database models. You would edit that file so it contains [DEFAULT] database_url = "https://alphafold.ebi.ac.uk/files/AF-{uniprot_id}-F1-model_v{version}.cif" last_update_time = 1e12 but replace the https url with your local URL pattern. Your local URL needs to have the {uniprot_id} and {version} where the Uniprot ID and the version number (e.g. 4) will be substituted in. The last_update_time line is time in seconds since 1970 and here is set to be a large number so that ChimeraX will never update these settings. Normally ChimeraX contacts our RBVI server periodically to check if the AlphaFold URL has changed so it can provide the new URL. That is to handle the situation where EBI changes the URL. The location of the alphafold_database-1 file is ~/Library/Application Support/ChimeraX/ on Mac ~/.local/share/ChimeraX/ on Linux C:\Users\<user-name>\AppData\Local\UCSF\ChimeraX\ on Windows as described in the ChimeraX documentation here https://www.cgl.ucsf.edu/chimerax/docs/user/preferences.html If you would like to see the AlphaFold settings ChimeraX code it is here on Github here https://github.com/RBVI/ChimeraX/blob/develop/src/bundles/alphafold/src/data... The Fetch button on the AlphaFold user interface in ChimeraX (menu Tools / Structure Prediction / AlphaFold) tries to fetch the file for the UniProt ID associated with the selected protein chain. If the selected chains is from a PDB file then it usually has the uniprot id in the file and ChimeraX will use that. But if the Uniprot ID is not available the Fetch button will try the ChimeraX k-mer search web service and that will fail for you. Tom
participants (2)
-
Jim Thorpe -
Tom Goddard