Ceilidh ... Re^6: Constrained Optimization

PC GAMESS/Firefly-related discussion club

Re^6: Constrained Optimization

Dear Alex,

Richard said exactly what I wanted to say.

Heavy atom positions are accurately determined from X-ray crystallography. But H positions cannot be determined due to its size. Neutron diffraction can provide accurate H positions but it is not easily feasible. Hence for all NMR, EFG or any property calculation using X-ray geometry it is very important to find accurate H positions without modifying heavy atom positions.

It would be wonderful if you could send me a custom build with extended dimensions.

Also like Richard said please consider including this to the next regular version of Firefly.

Thanks very much.
Regards
Damodaran

On Mon Apr 20 '09 7:39pm, Alex Granovsky wrote
----------------------------------------------
>Dear Damodaran,

>indeed, there are no more than 16 atoms that can be frozen using IFRZAT array.
>This is not an intrinsic limitation, it is easy to recompile and
>extend this limit to e.g., 100 or more atoms. Internally, this is
>related with the number of primitives that can be frozen using
>ifzmat array.

>Please let me know if you need custom build with extended dimensions.

>However, my question is related with the quality of the X-ray data you are using - what is the uncertainty in coordinates?
>Is it small enough to justify freezing of coordinates of all heavy atoms?

>Regards,
>Alex

>On Mon Apr 20 '09 7:07pm, Damodaran Krishnan wrote
>--------------------------------------------------
>>Dear Dr.Granovsky,

>>The manual says "Currently no more than 16 atoms may be frozen". Is it still true?

>>My system has about 65 heavy atoms and ~ 75 hydrogens. I would like to freeze 70 heavy atoms. I guess it is not possible with IFRZAT right?

>>Thanks
>>Damodaran

>>On Mon Apr 20 '09 4:59pm, Alex Granovsky wrote
>>----------------------------------------------
>>>Hi,

>>>you can also use IFRZAT array of $zmat group:

>>>Regards,
>>>Alex Granovsky

>>>On Fri Apr 17 '09 5:56pm, Damodaran Krishnan wrote
>>>--------------------------------------------------
>>>>Thanks very much Richard. Yes, it would be great if we could just specify the symbol or even atom #.

>>>>Damodaran

>>>>On Fri Apr 17 '09 3:57pm, Richard wrote
>>>>---------------------------------------
>>>>>Attached are two ways of doing this for a simple molecule (ethanol).

>>>>>If the input format ever gets dragged into the 21st century, it would be nice to be able to do this by putting some symbol meaning 'don't move this atom' in the coordinate block somewhere.

>>>>>Richard
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>On Thu Apr 16 '09 1:06am, Damodaran Krishnan wrote
>>>>>--------------------------------------------------
>>>>>>Dear Firefly users,

>>>>>>X-ray structures generally does not have hydrogen coordinates, hence I add H atoms to the structure and do a constrained optimization (i.e. optimize the H atom positions keeping heavy atoms fixed). This is easy to do with G03 since it lets me freeze XYZ coordinates of any atom. But I am a PCGAMESS/Firefly fan and also this is much faster, so could somebody suggest an easy way to fix coordinates of all heavy atoms and optimize all H atom positions using Firefly.

>>>>>>Firefly manual says “You cannot freeze Cartesian cords”

>>>>>>Thanks very much.

Tue Apr 21 '09 7:08pm

This message read 1477 times