Dear Scott, Fabulous question. I think a lot of people are thinking about this quite deeply right now but our experience with single acylation and charge ladder work (unpublished) was that even a single amino acid charge change had a profound impact on crystallization outcome in the 1,536 conditions we probe. We use microbatch currently which is more static than vapor diffusion where the dynamics add another layer of complexity to the process. It could well be a beautiful application of ML/AI and there is a lot of interest in this here but even our database (~20 million images and a diverse chemical and protein landscape) may not be large enough for the learning process.
We had good experience in the ML/AI area automating image analysis and now routinely using this (with help from Google Brian - https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0198883) but it required data from multiple generous sources in the community – both academic and industry. (Unapologetic and shameless plug for the National Crystallization Center (get a crystal) at http://getacrystal.org where we use this). I would love to add any good links in this area to our structural biology resources page - Structural Biology Resources | Hauptman-Woodward Medical Research Institute (buffalo.edu)<https://hwi.buffalo.edu/structural-biology-resources/>. Best, Eddie Edward Snell Ph.D. President and CEO | Hauptman-Woodward Medical Research Institute Director | NSF BioXFEL Science and Technology Center Professor, Materials Design and Innovation | University at Buffalo, SUNY p: +1 716 898 8631 | f: +1 716 898 8660 e: esn...@hwi.buffalo.edu<mailto:esn...@hwi.buffalo.edu> skype: eddie.snell Hauptman-Woodward Medical Research Institute 700 Ellicott Street | Buffalo, NY 14203-1102 hwi.buffalo.edu<https://hwi.buffalo.edu/> [hwi-logo-primary-horizontal] From: CCP4 bulletin board <CCP4BB@JISCMAIL.AC.UK> On Behalf Of Scott Classen Sent: Monday, April 4, 2022 3:07 PM To: CCP4BB@JISCMAIL.AC.UK Subject: [ccp4bb] Has anyone successfully used RoseTTAFold or AF2 to guide crystallization? Hello CCP4, Has anyone successfully used the available ML/AI protein folding tools to guide crystallization construct design? Maybe you had a protein or domain that was resistant to crystallization ef Warning! This message was sent from outside your organization and we were unable to verify the sender. sophospsmartbannerend Hello CCP4, Has anyone successfully used the available ML/AI protein folding tools to guide crystallization construct design? Maybe you had a protein or domain that was resistant to crystallization efforts and the folding algorithms predicted some loops or termini that were disordered? Then you trimmed or modified them in some way to aid in crystallization? Or if you haven’t done this yourself, are you aware of anyone who has? Thanks, Scott ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Scott Classen, Ph.D. ALS-ENABLE TomAlberTron Beamline 8.3.1 SIBYLS Beamline 12.3.1 Advanced Light Source Lawrence Berkeley National Laboratory 1 Cyclotron Rd MS6R2100 Berkeley, CA 94720 mobile 510.206.4418 desk 510.495.2697 beamline 510.495.2134 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ________________________________ To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCP4BB&A=1 ######################################################################## To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCP4BB&A=1 This message was issued to members of www.jiscmail.ac.uk/CCP4BB, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
