Hi Paul, You asked me to expand, let’s hope this is not overdone. I hesitated to send this to the entire list, but finally there might be useful information for those interested.
I was talking about the Moorhen embedded in CCP4. And you are right, I started using the web version with Brave and it works much faster than with my other browsers (had tried a few). I spent a good part of a day on a poor molecular replacement solution at ~2.7 angstroms, many copies in the AU. It was a nice test drive, even though I couldn’t try all the functions yet. I started to compile lists (not exhaustive yet, so I can also send you this later, if at all useful, when I accumulate more points): Things that I’m used to in Coot, couldn’t find them in Moorhen: - LINK two atoms: couldn’t find this in Coot-1 either… I’m afraid with the change to mmcif format, this will change a lot as well? LINK is very useful when we are refining coordinated metals, even a few very close bonds as well (some ADP-Mg-vanadate triads)… Are LINKs being replaced? I haven’t tested if the LINK records already present in the pdb file are still being taken in account, I hope they are. - Validate -> Probe Clashes: very fast way to add hydrogens in Coot 0.9xxx, and obviously to show clashes and other issues. - “Sphere refine” as in Coot-1. Please. - In the old Coot I could add the expected sequence of the entire ORF and go checking as I go, also helpful for missing loops and termini. All right, maybe not the most popular and essential function… Keyboard shortcuts would be nice to have in Moorhen: - “Drag atoms”. Because “refine triplet” already has one, good idea. - "Accept changes" (the old "Hit Enter" in Coot) after the drag atoms job. Problems I’ve found (Moorhen running with Brave, iMac 16 cores, Sequoia): - “Drag atoms” and “refine” seem to ignore or give lower weight to the geometry at the boundaries of the region being refined. I don’t remember having to go back to fix the boundaries so often with Coot. Am I doing something wrong? - On map contour: contour sliders, the +/- buttons and Ctrl+Scroll all change the contour level too fast! (Steps could be shorter, maybe?) - When I hit "Space", moving to next residue includes a spinning and zooming-in that looks very nice, but also makes me dizzy after a few residues. It even reverses the movement every 3 hits! I can’t do this after lunch. The most disturbing is the (re)zooming-in at every step, maybe we could have an option to go ahead without zooming. Could be a vintage version for older people like me. - The sequence viewer is very useful, just not very easy to learn: you need to drag to the right to go to the left... and drag the edge of the pan viewer to zoom in an out, it took me a while to figure that out (maybe the younger ones won’t suffer as much). And even with minimum zoom, I find the letters could still be smaller, to give more space for the structure. - On "Drag atoms" (my favorite function, thanks for that), picking is reluctant sometimes, I have to insist until it understands I want to pick an atom, and not rotate the view – even when I’m absolutely sure to be clicking on an atom position, not on the bond. - Less problematic: the ‘pop-ups’ can’t go out of the browser window… for those who were used to pushing them to the secondary screen. And yes, the interface is modern and well-conceived. I had 0 crashes so far. Because my Coot-1 is not working, I preferred to give Moorhen a go, and I’m happy with it. Moving and refinement are still slower than with the Coots, but the clicking routine is optimized, it pays off somehow, IMO (you are absolutely right, it doesn’t need to mimic Coot). Also, the map being updated as we fix the model (long time waited function, wasn’t working for me before!) is fantastic. I still need to try Moorhen on a structure with a fancy ligand and coordinated Mg2+, we’ve got a few of them coming soon, I hope. Just tried the “create from SMILES” function on something else, works very nicely- does it also run 100% locally? Asking for a friend who is paranoid about confidentiality. I’ll be willing for more control over the .cif restraints, though, in some of our very special cases. Or we’ll just keep Cooting those. Concerning my “old” iMac… it is actually a server version with boosted RAM, I still like it. I bought it in a hurry exactly to avoid the change of CPU architecture (afraid of what it would do to the xtal software back then). We’ll bother Santa with the cryoEM stuff… Anyways, thanks for all these developments, I wish all the best to you and all the contributors of Coot – not even sure how many you are, now. Cheers! --------------------------------------------------- Carlos KIKUTI, PhD UMR144 - CNRS - Institut Curie Pavillon Trouillet Rossignol 26 Rue d’Ulm - 75005 Paris, France carlos.kik...@curie.fr<mailto:carlos.kik...@curie.fr> De : Paul Emsley <pems...@mrc-lmb.cam.ac.uk> Date : mardi, 26 novembre 2024 à 10:26 À : Kikuti Carlos <carlos.kik...@curie.fr>, COOT@JISCMAIL.AC.UK <COOT@JISCMAIL.AC.UK> Objet : Re: Coot on OSX 15.1.1 On 25/11/2024 11:50, Kikuti Carlos wrote: -- I also upgraded to Sequoia ~ 10 days ago, Coot 0.9XXX (the one in the current CCP4 9 package) is working, not even needed to reinstall Xquartz. OK, good. Not sure this makes any difference: I had installed GTK4 through homebrew before installing CCP4 9 with the two Coots + Moorhen embedded. I doubt that that will make a difference to the Coot built/installed by CCP4. And I’m running one of the last Intel iMacs ever made. Maybe Santa will bring you and M4 Pro for Christmas (I hope he does for me). Coot-1 was sort of working before the upgrade (many issues with different windows, a few sudden crashes) and now it opens, but doesn’t do anything after that. I am in the same situation ATM. Hopefully resolved soon - but I don't know what the problem is. I think I’ll need to extend my patience a bit further… from what I could do before the upgrade, Coot-1 was really nicer and faster. Yes. Moorhen doesn’t do everything I need, Oh? Expand on this please. or maybe I didn’t find the options, Quite Possible. The Moorhen gui is/was designed to be a modern web application, not to mimic Coot. in any case it is way slower than the two Coots. Sad. I found Moorhen smooth to use in Chrome and other Chrome-based browsers (e.g. Brave). On Firefox, it was not so hot. For actual calculations, it is around 50-100% slower than Coot, but still amazing given that it's computing in the web browser. Anyways, just to let you know that I’m looking forward to instructions to make Coot-1 work as well, whenever they are available. And thank you Bill for the long time efforts! seconded. Paul. ######################################################################## To unsubscribe from the COOT list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=COOT&A=1 This message was issued to members of www.jiscmail.ac.uk/COOT, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
