Dear All, for a recent dataset, I would like to try MR-Rosetta, as normal MR didn't work (not surprising perhaps, as the closest model is from NMR and with only about 20% sequence identity). Even though I don't think there is much chance of success, I would like to try MR-Rosetta anyway, although without having to install Rosetta if possible. Is there a server that would take care of the whole process? I.e. take the input mtz and the hopefully homologous structure supplied by me, and run Rosetta followed by MR in Phenix? In its defect, is there an online server that does the Rosetta part of MR-Rosetta, that would give a Rosetta-improved output model for putting into MR by Phenix, which I already have installed? Some side-ways related questions: - is it feasible to install Rosetta under MacOSX for a not very computer-literate user? - Phenix seems to slow down my MacPro a lot, even if I use just one of the 8 processors - is 4 Gb of RAM too little? Sorry for the rather rambling mail... Greetings, Mark Mark J van Raaij Lab 20B Dpto de Estructura de Macromoleculas Centro Nacional de Biotecnologia - CSIC c/Darwin 3 E-28049 Madrid, Spain tel. (+34) 91 585 4616 http://www.cnb.csic.es/~mjvanraaij
Hi Mark, I don't know of any server for mr_rosetta at present, but the Robetta server can do the homology modeling part: http://robetta.bakerlab.org/faqs.jsp#whatis_robetta It is not too difficult to install rosetta; when I do it I just cut and paste and edit the instructions that are on the phenix mr_rosetta web page at https://www.phenix-online.org/documentation/mr_rosetta.htm#anch35 (Installing Rosetta for use with mr_rosetta). All the best, Tom T ________________________________________ From: [email protected] [[email protected]] on behalf of Mark J van Raaij [[email protected]] Sent: Tuesday, April 02, 2013 12:00 PM To: [email protected] Subject: [phenixbb] MR-Rosetta server? Dear All, for a recent dataset, I would like to try MR-Rosetta, as normal MR didn't work (not surprising perhaps, as the closest model is from NMR and with only about 20% sequence identity). Even though I don't think there is much chance of success, I would like to try MR-Rosetta anyway, although without having to install Rosetta if possible. Is there a server that would take care of the whole process? I.e. take the input mtz and the hopefully homologous structure supplied by me, and run Rosetta followed by MR in Phenix? In its defect, is there an online server that does the Rosetta part of MR-Rosetta, that would give a Rosetta-improved output model for putting into MR by Phenix, which I already have installed? Some side-ways related questions: - is it feasible to install Rosetta under MacOSX for a not very computer-literate user? - Phenix seems to slow down my MacPro a lot, even if I use just one of the 8 processors - is 4 Gb of RAM too little? Sorry for the rather rambling mail... Greetings, Mark Mark J van Raaij Lab 20B Dpto de Estructura de Macromoleculas Centro Nacional de Biotecnologia - CSIC c/Darwin 3 E-28049 Madrid, Spain tel. (+34) 91 585 4616 http://www.cnb.csic.es/~mjvanraaij _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb
On Tue, Apr 2, 2013 at 11:00 AM, Mark J van Raaij
- is it feasible to install Rosetta under MacOSX for a not very computer-literate user?
Yes, but if you've never had to install software from source before it may be a little intimidating. (Not sure if they plan to release binary builds at any point - will check on this.)
- Phenix seems to slow down my MacPro a lot, even if I use just one of the 8 processors - is 4 Gb of RAM too little?
In general it is wise to have at least 1 GB per core, and preferably 2+ GB. The slowdown is probably due to other applications swapping to disk to make room for Phenix; adding memory will definitely help with this. The amount you actually need depends on what kind of structures you're working on - but keep in mind that other apps such as web browsers are also memory hogs. (Watching YouTube while refining structures on the same computer is probably not going to work very well with only 4GB.) -Nat
participants (3)
-
Mark J van Raaij -
Nathaniel Echols -
Terwilliger, Thomas C