[phenixbb] Simulated annealing omit maps

[Thomas C. Terwilliger]

Dear Chandrasekhar,

The 'number of boxes' is the available way to define how small the portion
of the map (and model) to omit in each piece of the composite omit map.

You could estimate how many boxes to use from your solvent content and the
fraction of the model you want to omit (50% solvent and 5% of model
omitted would mean 2.5% of the a.u to be omitted each box, or 40 boxes.)

I haven't systematically analyzed the best box size, and would be
interested to hear of others' experiences with this.

All the best,
Tom T

  >> Hello friends,
>>
>> I have a couple of questions regarding simulated annealing composite map
>> calculations.
>> (1)  Is there a way of specifying the percentage of residue co-ordinates
>> to be omitted each time as we move through the asymmetric unit?
>> I have found, when using CNS earlier, that depending on the structure,
>> the map quality often varies if you change this from the usual 5% to 4%
>> or even 3%, sometimes getting clearer.
>>
>> (2) Does the 'number of boxes' you can specify in the GUI only affect the
>> fine structure/grid spacing of the map?  The map probably does not get
>> better all the time as we keep on increasing the number of boxes; when
>> do we reach the point of diminishing returns and what would be the best
>> way to estimate it?
>>
>> Thank you very much and wish you all a Very Happy 2012.
>>
>> K. Chandrasekhar
>>
>>
>> _______________________________________________
>> phenixbb mailing list
>> phenixbb at phenix-online.org
>> http://phenix-online.org/mailman/listinfo/phenixbb
>>