From: Yarrow Madrona
Date: 6 December 2011 16:03
Hello,
I am using scalepackvirus and I noticed that the rejection list grows but
does not disappear in later rounds of scaling from the log file as in
scalepack. I am assuming that the rejections are treated the same way as
in scalepack but for some reason are not removed from the log file. Does
anyone know if this is correct?
Also I wondered if there is any way to get a redundancy dependent R value
from scalepack. Thank you.
-Yarrow
--
Yarrow Madrona
----------
From: Schubert, Carsten [JRDUS]
Yarrow,
as far as I know, scalepack, scalepackvirus and the other variants only
differ by their respective array sizes (Someone from Wladek's or ZO's
lab may be more suitable to comment on this). The rejections in the log
file are controlled by the parameters 'write rejection file xxx' and
'rejection probability'. You may want to read up on those in the manual.
I believe that historically the way it was supposed to work is that
rejection candidates are written to the log file for examination and
then in subsequent rounds are written to the 'reject' file. The latter
is determined by the 'write rejection file' parameter.
No, scalepack does not output the rmeas or rpim values. I have some
recollection that Manfred Weiss at EMBL wrote a program which could
analyze scalepack data written with 'no merge original index' to yield
those parameters, you may want to look into this. The name of the
fortran source is 'rmerge.f'. Unfortunately the link to the original
source repository is dead, but I have a copy around, which I could
share.
Another way would be to go the route of scalepack unmerged data ->
pointless -> scala, although I have never tried that myself.
Good luck
Carsten
----------
From: Jim Pflugrath
Folks may wish to watch a webinar I gave on using HKL3000. In the webinar I describe how to calculate an Rmeas from scalepack output and also discuss a few other "how-to" operations.
See www.rigaku.com/protein/webinars-past.html for more viewing pleasure.
Jim
----------
From: Yarrow Madrona
Thank you for your help,
I had the write rejection file set to 0.5 and probability at 0.001. It
seems that I had so many rejections that it was hard to get rid of all of
them. I have obtained a script to keep running scalepack until all of the
rejections are gone from the log file. It takes a very long time to run
but it is better than hitting enter over and over.
Although I have read the manual I am still having some difficulty
understanding the difference between rejection file and the probability.
If rejection file is set to 0.5 I'm guessing that if a reflection has a
50% probability of being an outlier it will be written to the log file.
But I don't understand how the rejection probability interplays here.
Thanks for any help you can provide.
-Yarrow
Date: 6 December 2011 16:03
Hello,
I am using scalepackvirus and I noticed that the rejection list grows but
does not disappear in later rounds of scaling from the log file as in
scalepack. I am assuming that the rejections are treated the same way as
in scalepack but for some reason are not removed from the log file. Does
anyone know if this is correct?
Also I wondered if there is any way to get a redundancy dependent R value
from scalepack. Thank you.
-Yarrow
--
Yarrow Madrona
----------
From: Schubert, Carsten [JRDUS]
Yarrow,
as far as I know, scalepack, scalepackvirus and the other variants only
differ by their respective array sizes (Someone from Wladek's or ZO's
lab may be more suitable to comment on this). The rejections in the log
file are controlled by the parameters 'write rejection file xxx' and
'rejection probability'. You may want to read up on those in the manual.
I believe that historically the way it was supposed to work is that
rejection candidates are written to the log file for examination and
then in subsequent rounds are written to the 'reject' file. The latter
is determined by the 'write rejection file' parameter.
No, scalepack does not output the rmeas or rpim values. I have some
recollection that Manfred Weiss at EMBL wrote a program which could
analyze scalepack data written with 'no merge original index' to yield
those parameters, you may want to look into this. The name of the
fortran source is 'rmerge.f'. Unfortunately the link to the original
source repository is dead, but I have a copy around, which I could
share.
Another way would be to go the route of scalepack unmerged data ->
pointless -> scala, although I have never tried that myself.
Good luck
Carsten
----------
From: Jim Pflugrath
Folks may wish to watch a webinar I gave on using HKL3000. In the webinar I describe how to calculate an Rmeas from scalepack output and also discuss a few other "how-to" operations.
See www.rigaku.com/protein/webinars-past.html for more viewing pleasure.
Jim
----------
From: Yarrow Madrona
Thank you for your help,
I had the write rejection file set to 0.5 and probability at 0.001. It
seems that I had so many rejections that it was hard to get rid of all of
them. I have obtained a script to keep running scalepack until all of the
rejections are gone from the log file. It takes a very long time to run
but it is better than hitting enter over and over.
Although I have read the manual I am still having some difficulty
understanding the difference between rejection file and the probability.
If rejection file is set to 0.5 I'm guessing that if a reflection has a
50% probability of being an outlier it will be written to the log file.
But I don't understand how the rejection probability interplays here.
Thanks for any help you can provide.
-Yarrow
No comments:
Post a Comment