Re: [ccp4bb] Refienmnet

[Eleanor Dodson]

There is a REFMAC plot of Rfactor v resolution, and of  v  v
resolution.
These can show if there is some resolution dependent spike - e.g. one due
to ice rings or saturated low resin data.
Look at those I suggest
Eleanor

On 17 August 2017 at 18:38, Kay Diederichs 
wrote:

> Dear Rohit,
>
> in your refinement table I find that the R-factors are elevated in the ice
> ring ranges (around 2.22 and 1.93 A resolution).
> And your data collections statistics seem to confirm this -  CC1/2 is very
> high at high resolution which is often due to ice rings.
>
> Could be a classical case of "data collection problem becomes visible at
> refinement stage".
>
> best,
>
> Kay
>
> On Tue, 15 Aug 2017 22:09:44 +0530, rohit kumar 
> wrote:
>
> >Dear All,
> >
> >
> >
> >I am refining a data in refmac5 and the data resolution is 1.8 A. Data is
> >looking fine with the data statics see below for data statics
> >
> >
> >
> >[image: Inline image 1]
> >
> >
> >
> >Right now the R/Rfree is 22/26 not good for such resolution. I have tried
> >most of the options in refmec5 but still I am not able to lower these
> >R/Rfree values.
> >
> >During refinement with phenix what I found that at resolution 1.93 to 1.86
> >R/Rfree is quite high see below
> >
> >
> >[image: Inline image 2]
> >
> >
> >
> >
> >Is this the possible reason for high R/Rfree value?. If it is  please let
> >me know how some can remove these frames during refinement. or
> >
> >
> >Please tell me other strategies to lower down the R/Rfree values.
> >
> >
> >Thank you in advance
> >
> >
> >With Regards
> >
> >Dr. Rohit Kumar Singh
> >
>

Re: [ccp4bb] Refienmnet

[Kay Diederichs]

Dear Rohit,

in your refinement table I find that the R-factors are elevated in the ice ring 
ranges (around 2.22 and 1.93 A resolution).
And your data collections statistics seem to confirm this -  CC1/2 is very high 
at high resolution which is often due to ice rings.

Could be a classical case of "data collection problem becomes visible at 
refinement stage".

best,

Kay

On Tue, 15 Aug 2017 22:09:44 +0530, rohit kumar  wrote:

>Dear All,
>
>
>
>I am refining a data in refmac5 and the data resolution is 1.8 A. Data is
>looking fine with the data statics see below for data statics
>
>
>
>[image: Inline image 1]
>
>
>
>Right now the R/Rfree is 22/26 not good for such resolution. I have tried
>most of the options in refmec5 but still I am not able to lower these
>R/Rfree values.
>
>During refinement with phenix what I found that at resolution 1.93 to 1.86
>R/Rfree is quite high see below
>
>
>[image: Inline image 2]
>
>
>
>
>Is this the possible reason for high R/Rfree value?. If it is  please let
>me know how some can remove these frames during refinement. or
>
>
>Please tell me other strategies to lower down the R/Rfree values.
>
>
>Thank you in advance
>
>
>With Regards
>
>Dr. Rohit Kumar Singh
>

Re: [ccp4bb] Refienmnet

[Dale Tronrud]

   Your first step is to look at your images and see what is going on in
that shell.  Since you are looking at merged stats the first guess is
that there is something wrong with all of the images, but only by
looking at them can you tell.

Dale Tronrud

On 8/15/2017 9:39 AM, rohit kumar wrote:
> Dear All,
> 
>  
> 
> I am refining a data in refmac5 and the data resolution is 1.8 A. Data
> is looking fine with the data statics see below for data statics
> 
>  
> 
> Inline image 1
> 
>  
> 
> Right now the R/Rfree is 22/26 not good for such resolution. I have
> tried most of the options in refmec5 but still I am not able to lower
> these R/Rfree values. 
> 
> During refinement with phenix what I found that at resolution 1.93 to
> 1.86 R/Rfree is quite high see below 
> 
> 
> Inline image 2
> 
> 
> 
> 
> Is this the possible reason for high R/Rfree value?. If it is  please
> let me know how some can remove these frames during refinement. or
> 
> 
> Please tell me other strategies to lower down the R/Rfree values.  
> 
> 
> Thank you in advance
> 
> 
> With Regards
> 
> Dr. Rohit Kumar Singh
> 
>