[phenixbb] reflection outliers detection in phenix.refine

[guanrj at foxmail.com]

Thank you, Pavel. 

I am not clear about "model_based_outliers"; I searched phenix.refine documentation 
but did not find anything. Is it based on my structure model?
I will look at the paper you suggested.    

By the way, "main.outliers_rejection=False" can also keep those model_based_outliers?
http://www.phenix-online.org/documentation/refinement.htm did not mention this type
of outliers:
outliers_rejection= True Remove basic wilson outliers , extreme wilson
                          outliers , and beamstop shadow outliers

Thanks

Rongjin 

= = = = = = = = On 2010-12-08 15:28:59 You wrote = = = = = = = = 
Rongjin,

outliers detection and rejection was inspired by this paper
R.Read, Acta Cryst. (1999). D55, 1759-1764.

and the paper by R.Agarwal's (197?/198?) mentions this too.

You can always turn this off using "main.outliers_rejection=False" if you believe it's necessary.

It's probably the quality of your data that causes the amount of outliers you quote.

Pavel.

PS> FYI: I got your email sent off-list and I always get the emails sent to phenixbb, so there is no need to duplicate them. 


On 12/8/10 11:45 AM, guanrj at foxmail.com wrote: 
Hi All

I have a question regarding reflection outliers detection in phenix.refine:
in my input reflection file, there are 24432 reflections, but only 22769 of them 
were used in refinement. I checked the log file and saw 1663 outliers. 

(below if from the log)
basic_wilson_outliers    = 14
extreme_wilson_outliers  = 18
beamstop_shadow_outliers = 0
model_based_outliers     = 1657
total                    = 1663

What is the  "model_based_outliers"?  What causes so many reflections rejected
and not used in the refinement?     

Thanks

Rongjin 


_______________________________________________
phenixbb mailing list
phenixbb at phenix-online.org
http://phenix-online.org/mailman/listinfo/phenixbb
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phenix-online.org/pipermail/phenixbb/attachments/20101208/0f75e786/attachment-0003.htm>