Phenix: Resources
Feb 16, 2010
Phenix (Python-based Hierarchical ENvironment for Integrated Xtallography) has been developed for the determination of macromolecular crystallographic structures. Phenix is a leader in automating the structure determination process.
I am going to start playing around with Phenix, but first wanted bring together helpful resources about the program suite:
- Documentation (479 pg pdf version)
- Publication list
- Phenix User Mailing List (Archives)
- CCP4bb (search results)
- 2007 Presentation by Paul Adams
- CCP4 Wiki on Phenix
Are you using Phenix? Why or why not?
What resources have you found especially helpful? Do you have any tips or tricks for getting started?
Posted by Sean |
Categories: Uncategorized | Tagged: Learning, Macromolecular Crystallography, Phenix |
Kat
February 17th, 2010 at 12:51 PM #
I started using Phenix about a year ago when I collected some data that was in that weird resolution cusp where you have too much data for CNS and not enough for SHELX. The automatic NCS was also incredibly attractive. I am definitely addicted now. I can’t imagine life without it.
The online manual is awesome. It explains everything very clearly and if you run into a problem, odds are someone already asked about it on the bulletin board. Like the COOT and CCP4 bulletin boards, the Phenix developers are quick to address any questions or issues. It would be nice if the bb archive had a search function, but I’ve found that a google “in site” search works just as well.
I’d suggest learning to use it in the command line first as it shows you the breadth of options you have available to you, like optimizing the mask and weights (wxu and wxc). You have to search these options out in the GUI and unless you know they are there it is easy to overlook them.
AaronB
February 17th, 2010 at 11:37 PM #
I am also a big fan of Phenix. I started playing around with it late last year, and it really has given me a lot of headway with my structure. I typically find information on the BB, the developers are very friendly and prompt! I agree with Kat, command line is the way to go. The first time i used the gui, it was trying to place waters in my model, which is nowhere near that stage yet.
Like a lot of these programs, make sure you have a fairly powerful computer. I ran a SA omit map last week, it took nearly 4 days on a dual core system! (my structure also has about 12 chains, so that does not help) Btw, if you plan on using the SOLVE, RESOLVE algorithms my suggestion is to have tons of ram.
chempilot
February 18th, 2010 at 11:53 AM #
bulletin board:
http://phenix-online.org/pipermail/phenixbb/
these guys are great. fast to respond and go way out of there way to help…
as a crystal structure newb, i have no idea if it’s lacking in anything. i find in that i constantly have to fight against a lot of assumed knowledge (i.e. crystallography jargon, statistics, command line speak, linux-speak, etc.) and a lack of just basic practical knowledge in books or the web (that’s why i liked your excellent beamline centering demo). so, i use PHENIX because of its ease of use. it seems to have clearer explanations of what it’s doing than CCP4, CNS or HKL. as i’ve learned more, i’ve slowly started poking around CCP4 and have begun to use command line entries, but i find my self going back to PHENIX to learn what i’m doing…
Sean
February 19th, 2010 at 10:25 AM #
@Kat and AaronB
Thanks so much for sharing your guidance. I will start off with the command line approach as you both suggested.
@chempilot
Funny. I was initially discouraged from using Phenix because it was so automated. The worry being that one would develop only a ‘black box’ approach to understanding crystallographic refinement. I appreciate how you are coming at the process ‘I can get this to work’ that then gives you a reference point to learn other software.
DrSNO
February 21st, 2010 at 8:44 PM #
The problem I have had with phenix is that it runs very slowly, especially compared to ccp4. However, the newest version is better and runs faster. I do like phenix refine, and run that seperately. Works fast, easy and is a great alternative to refmac. It’s interaction with Richardson’s kinemag and coot is great.