Unusually low B factors
Hi all, I have produced xenon derivatives for SAD phasing of several standard proteins by placing them under xenon gas pressure and then immediately flash freezing in liquid nitrogen. Everything has worked out nicely as far as successful phasing, nice models, and low R factors. Now I've noticed that the B values have decreased from initial processing values. Several rounds of manual changes and phenix.refine have caused B values to decrease each round of refinement to values of around 10 (I guess 20 is a more common value). Does anyone know why this is happening or what the explanation might be? I'm not sure what other information to include so please contact to discuss further. Thanks! ~Brennan
On Wed, Apr 17, 2013 at 12:19 PM, Brennan Bonnet
I have produced xenon derivatives for SAD phasing of several standard proteins by placing them under xenon gas pressure and then immediately flash freezing in liquid nitrogen. Everything has worked out nicely as far as successful phasing, nice models, and low R factors.
Now I've noticed that the B values have decreased from initial processing values. Several rounds of manual changes and phenix.refine have caused B values to decrease each round of refinement to values of around 10 (I guess 20 is a more common value). Does anyone know why this is happening or what the explanation might be?
Which version are you using? If it's 1.8.2 or earlier, could you please update to the latest nightly build and try again? -Nat
Hi Nat,
I'm running version 1.8.2-1309.
I could try this again if I need the newer version. I'm currently completing my thesis and have processed and reprocessed this data so many times that. At this point it's getting more important that I wrap up my degree than getting this processing to work 100%.
I'm hoping that someone else has had this problem with B values before and could offer an explanation like "It's due to gas pressure, or sometimes B values are just low and that's acceptable". That way I can include it in my thesis so that I don't get grilled by my committee.
~Brennan
________________________________________
From: [email protected] [[email protected]] On Behalf Of Nathaniel Echols [[email protected]]
Sent: Wednesday, April 17, 2013 1:23 PM
To: PHENIX user mailing list
Subject: Re: [phenixbb] Unusually low B factors
On Wed, Apr 17, 2013 at 12:19 PM, Brennan Bonnet
I have produced xenon derivatives for SAD phasing of several standard proteins by placing them under xenon gas pressure and then immediately flash freezing in liquid nitrogen. Everything has worked out nicely as far as successful phasing, nice models, and low R factors.
Now I've noticed that the B values have decreased from initial processing values. Several rounds of manual changes and phenix.refine have caused B values to decrease each round of refinement to values of around 10 (I guess 20 is a more common value). Does anyone know why this is happening or what the explanation might be?
Which version are you using? If it's 1.8.2 or earlier, could you please update to the latest nightly build and try again? -Nat _______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb
On Wed, Apr 17, 2013 at 12:41 PM, Brennan Bonnet
I could try this again if I need the newer version. I'm currently completing my thesis and have processed and reprocessed this data so many times that. At this point it's getting more important that I wrap up my degree than getting this processing to work 100%.
I'm hoping that someone else has had this problem with B values before and could offer an explanation like "It's due to gas pressure, or sometimes B values are just low and that's acceptable". That way I can include it in my thesis so that I don't get grilled by my committee.
Well, there is natural variability in the B-factors at any resolution, and no direct equivalence to the Wilson B-factor calculated from intensity statistics. But the behavior you are seeing is an implementation detail of some versions of Phenix - it all depends on whether the overall scale is incorporated into the atomic B-factors. Because most refinement programs (including most versions of Phenix aside from a few of the 1.8 series) always do this, most crystallographers have come to expect a specific range for B-factors at any given resolution. If you want to be absolutely certain that your committee is happy, just re-run the last round of refinements with the nightly build version - no need to do any re-processing or anything fancy. -Nat
I'm hoping that someone else has had this problem with B values before and could offer an explanation like "It's due to gas pressure, or sometimes B values are just low and that's acceptable". That way I can include it in my thesis so that I don't get grilled by my committee.
It's been discussed on this mailing list recently so many times so I am super surprised that you did not notice it (assuming that you read emails on this list). If it is what we think it is then it is an unwanted behavior of the program (not a bug, just an unwanted behavior) that is fixed by now and so the suggested is to use a recent version. Pavel
Hi Brennan, using recent Phenix will fix the problem, let me know otherwise.
phenix.refine have caused B values to decrease each round of refinement to values of around 10 (I guess 20 is a more common value).
B~20A**2 is a more common value if data resolution is ~1.25...1.75A. Otherwise a more common value can be very different from 20A**2. See plot below. Pavel
participants (3)
-
Brennan Bonnet -
Nathaniel Echols -
Pavel Afonine