![](https://secure.gravatar.com/avatar/1b74f8b9dffef7e9fce75bfc7e23ffd1.jpg?s=120&d=mm&r=g)
Hi Frank, "such a tool" is at http://strucbio.biologie.uni-konstanz.de/xdswiki/index.php/Aniso_cutoff where it's meant to be applied to INTEGRATE.HKL which comes out of XDS. Doing it this way has the benefit that the statistics printed out by XDS' CORRECT (or SCALA/TRUNCATE; there are people who prefer that route) match the data you refine agains. HTH, Kay
Message: 1 Date: Thu, 15 Jul 2010 06:32:40 +0100 From: Frank von Delft
To: PHENIX user mailing list Subject: [phenixbb] Selecting ellipsoid of data Message-ID:<[email protected]> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Hi, is there a tool in phenix that allows me to select an ellipsoid of data -- specified e.g. by the highest resolutions in three reciprocal lattice directions. (Yes, I'm playing with anisotropy, "playing" being the operative word.)
phx
------------------------------
Message: 2 Date: Wed, 14 Jul 2010 22:47:30 -0700 From: "Ralf W. Grosse-Kunstleve"
To: [email protected] Subject: Re: [phenixbb] Selecting ellipsoid of data Message-ID:<[email protected]> Content-Type: text/plain; charset=us-ascii Hi Frank,
Hi, is there a tool in phenix that allows me to select an ellipsoid of data -- specified e.g. by the highest resolutions in three reciprocal lattice directions. (Yes, I'm playing with anisotropy, "playing" being the operative word.)
I'm not aware of such a tool.
Ralf
------------------------------
Message: 3 Date: Thu, 15 Jul 2010 07:15:58 +0100 From: Frank von Delft
To: "Ralf W. Grosse-Kunstleve" , PHENIX user mailing list Subject: Re: [phenixbb] Selecting ellipsoid of data Message-ID:<[email protected]> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Hi Ralf
Yeah, I figured. So if I want to use cctbx, where do I start? Just a pointer to a) package and b) function where I'll see the syntax.
So equation for ellipsoid is x^2/a^2 + y^2/b^2 + z^2/c^2 = 1; so I imagine I take each reflection, convert each of h,k,l to 1/reso, and with a = 1/res(a*), I just check whether the above is< 1.
The main thing I still need is to convert h,k,l into orthogonal coordinates.... or do I? I suppose I don't, as what I care for is not whether it's "really" an ellipsoid, only whether it cuts through miller index space anisotropically.
Hmmmm... I may be able to do it in sftools; but if you can in<1minute give me a link to where to look to get started in cctbx, that would be awesome.
(Thanks for listening :)
Hi Frank,
Hi, is there a tool in phenix that allows me to select an ellipsoid of data -- specified e.g. by the highest resolutions in three reciprocal lattice directions. (Yes, I'm playing with anisotropy, "playing" being the operative word.)
I'm not aware of such a tool.
Ralf