[ dev-564 on GNU/Linux ] why does phenix - apparently - keep putting the label "merged" in output .def's, when neither the input .def or .mtz has this label - and what generally is the significance? the .mtz has separate Bijvoets, and I's. e.g: input .def: (concise) labels = "I(+),SIGI(+),I(-),SIGI(-)" label = "R-free-flags" output .def: labels = "I(+),SIGI(+),I(-),SIGI(-),merged" label = "R-free-flags" i re-checked the .mtz lacks "merged" (not shown). -bryan
Hi Bryan, the "merged" appears only if the input mtz contains symmetry-equivalent hkl. phenix.refine merges symmetry-equivalent reflections to obtain exactly one reflection in the reciprocal-space asymmetric unit. (The merging algorithm is very simple; it is better to merge with a dedicated data reduction program.) Ralf ----- Original Message ----
From: Bryan Lepore
To: PHENIX user mailing list Sent: Fri, October 29, 2010 12:44:51 PM Subject: [phenixbb] refine : "merged" in .def [ dev-564 on GNU/Linux ]
why does phenix - apparently - keep putting the label "merged" in output .def's, when neither the input .def or .mtz has this label - and what generally is the significance?
the .mtz has separate Bijvoets, and I's. e.g:
input .def: (concise) labels = "I(+),SIGI(+),I(-),SIGI(-)" label = "R-free-flags"
output .def: labels = "I(+),SIGI(+),I(-),SIGI(-),merged" label = "R-free-flags"
i re-checked the .mtz lacks "merged" (not shown).
-bryan _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb
Hi Ralf,
On Fri, Oct 29, 2010 at 4:00 PM, Ralf W. Grosse-Kunstleve
the "merged" appears only if the input mtz contains symmetry->equivalent hkl. >phenix.refine merges symmetry-equivalent reflections to obtain >exactly one reflection in the reciprocal-space asymmetric unit. (The merging algorithm is very simple
... that makes sense - IIRC, phenix can write out all reflections/flag data again in a .mtz, so "merged" would show up then? -bryan
On Fri, Oct 29, 2010 at 1:07 PM, Bryan Lepore
the "merged" appears only if the input mtz contains symmetry->equivalent hkl. >phenix.refine merges symmetry-equivalent reflections to obtain >exactly one reflection in the reciprocal-space asymmetric unit. (The merging algorithm is very simple
... that makes sense - IIRC, phenix can write out all reflections/flag data again in a .mtz, so "merged" would show up then?
No, but the output data will indeed be merged (and possibly converted to amplitudes - it may depend on the program). PHENIX handles MTZ files much differently than CCP4 does, and there are further complications involving the use of other formats, so you can't assume to much about how the labels we report correspond to column labels in the MTZ file. -Nat
participants (3)
-
Bryan Lepore
-
Nathaniel Echols
-
Ralf W. Grosse-Kunstleve