[phenixbb] isomorphous Fobs-Fobs maps

Pavel Afonine pafonine at lbl.gov
Mon Jan 20 11:03:40 PST 2020


Hi Eike,

there isn't much documentation on this.. mostly because this is a rather 
simple tool..

Really, this is just a Fourier map {(Fobs_1 - Fobs_2), 
phase_from_given_model}. Here both, Fobs_1 and Fobs_2 are scaled to 
Fmodel as describe here:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3606040/pdf/d-69-00625.pdf

with overall scale k_total being applied back to Fobs: Fobs_new = 
Fobs/k_total. Technically this should bring both Fobs sets onto common 
scale.

The input model is assumed to fit both Fobs sets reasonably well, and it 
is not supposed to have the feature you are interested in to see in the 
Fo-Fo map.

I'm happy to investigate this more, but for that I will need more 
details (sent off list), such as:
- both Fobs data sets;
- model used as a source of phases;
- CCP4 generated map;
- a brief description of what you expect to see (and where) and why you 
believe what you get isn't consistent with what you expect.

Pavel

On 1/20/20 03:01, Schulz, Eike-Christian wrote:
> Dear all,
>
> we have recently tried to calculate isomorphous difference maps with phenix using the GUI. Unfortunately, we found that the results do _not_ seem to produce results that we would expect of a true difference map. Rather the resulting maps appear to show noise.
>
> We speculate that this is the result of _not_ scaling the datasets together.
>
> - Is this correct ?
>
> Applying the "multiscaling method" found in the advanced options changes the result, but it is still inconsistent with what we gain from e.g. the CCP4 procedures.
>
> Unfortunately, there is no online documentation to this tool.
>
> - Is there a way to scale the data to a reference before calculating the difference maps?
>
> - What is the proposed way of using this tool ?
>
> - Are there known bugs in the GUI version that are absent in the command line edition?
>
>
> With best regards,
>
> Eike
>   
>
>
> _______________________________________________
> phenixbb mailing list
> phenixbb at phenix-online.org
> http://phenix-online.org/mailman/listinfo/phenixbb
> Unsubscribe: phenixbb-leave at phenix-online.org



More information about the phenixbb mailing list