Re: [ccp4bb] Defining beamstop and error during indexing- moslfm

[Andrew Leslie]

Hi Sonali,

         How did you assign the spacegroup ... did you run POINTLESS ?
A table of Rmerge vs resln from SCALA would be helpful in addition to a
screen shot of an image as Francis suggested.

Andrew

> Dear Andrew and all the people for their help,
> I am providing mosflm the right beamstop and now, I am able to do the
> indexing, refinement and indexing too.Then I run scala for the output mtz
> file and it shows Rmerge too high 0.58and also when examining the spots
> and predictions in the image, it seems like it is not picking a lot of
> spots, so missing  large number of reflections.So, any suggestions to
> correct it or am i doing something wrong.
> Thanks again for the help and suggestions
>
> --
> Sonali Dhindwal
>
> “Live as if you were to die tomorrow. Learn as if you were to live
> forever.”
>
> --- On Thu, 22/3/12, Andrew Leslie <and...@mrc-lmb.cam.ac.uk> wrote:
>
> From: Andrew Leslie <and...@mrc-lmb.cam.ac.uk>
> Subject: Re: [ccp4bb] Defining beamstop and error during indexing- moslfm
> To: CCP4BB@JISCMAIL.AC.UK
> Date: Thursday, 22 March, 2012, 7:43 PM
>
> Dear Sonali,
>
>             Just to add to Tim's reply, when you open the image with
> iMosflm, you can Drag and drop the direct beam position in the
> image display window. First, you have to click on the leftmost
> icon in the row of icons under the image filename (a green
> cross) which will display the direct beam position as read
> from the image header as a green cross on the image. You can
> then drag then over to where you think the beam position
> should be (judging from the position of the backstop shadow).
>
> If indexing still does not work, you can try doing a direct beam search
> from the indexing pane, which will do a grid search +/- 2mm from the
> current position.
>
> However, you need to beware of one thing. I believe that the rotation axis
>  on the Sercat beamline goes in the opposite direction to most beam lines.
> In such cases, indexing using more than one image will never work, because
> the relative phi values will be wrong. You can try indexing with one
> image, and then if that works, try predicting the next image. If that
> prediction does not match, it is almost certainly because the rotation
> direction is wrong.
>
> To deal with this, choose the Settings->Experiment settings menu and click
> on the "reverse phi" box. Then you MUST repeat the spot search for each
> image being used in indexing, then it should all work.
>
> Best wishes,
>
> Andrew
>
>> Dear All,
>>
>> We have collected a data for a protein crystal at SER-CAT Chicago and
>> the
>> detector is mar300.We are using mosflm to process the data.While
>> indexing,  the beamstop which it is taking is wrong, due to which it
>> fails.
>>
>> I am trying to define the beamstop manually using tools like mask and
>> spot
>> search area.
>> (it might be wrong)
>> It will be highly appreciable if someone can please suggest if this the
>> method we should use to define the beamstop or there is any site
>>  definition file which has to be used, as it is available for HKL2000
>> or
>> how we have to define the beamstop in mosflm.
>>
>>
>>
>> --
>> Sonali Dhindwal
>>
>> “Live as if you were to die tomorrow. Learn as if you were to live
>> forever.”
>