[Biopython] SVDSuperimposer()

[George Devaniranjan]

Thank you João,

No, I am using an in-house code written in python but in that code I use
SVDSuperimposer() as well.
The "energy function" that is minimized has various terms such as a steric
clash is highly disfavored, Hydrogen bonds are favored...etc

Let me try your second suggestion and see if that helps.


On Fri, Apr 11, 2014 at 4:37 PM, João Rodrigues <anaryin at gmail.com> wrote:

> Hi George,
>
> Sorry for the delay in the answer..
>
> Are you doing the minimization using Biopython? That's the only way I see
> in which the SVDSuperimposer is a bottleneck. In any case, the SVD code is
> written in C, so it should be pretty fast. Can you identify precisely where
> the bottleneck is (atom selection, fitting, calculation, etc)?
>
> Anyway, I would suggest looking into some weak position restraints on the
> heavy atoms of the backbone to keep things sort of in place. This would
> avoid the RMSD calculations at every step (I guess?), instead just a simple
> harmonic potential calculation added to the energy function.
>
> Cheers,
>
> João
>
>
>
>
> 2014-04-11 18:22 GMT+02:00 George Devaniranjan <devaniranjan at gmail.com>:
>
> Oh, sorry-yes I meant the speed.
>>
>>
>> On Fri, Apr 11, 2014 at 12:11 PM, João Rodrigues <anaryin at gmail.com>wrote:
>>
>>> Hey George,
>>>
>>> What do you mean by bottleneck? In terms of speed?
>>>
>>> You can always use Profit for example to calculate RMSDs between the
>>> models. It's a bit faster than our module.
>>>
>>> Cheers,
>>>
>>> João
>>>
>>>
>>> 2014-04-11 17:58 GMT+02:00 George Devaniranjan <devaniranjan at gmail.com>:
>>>
>>>> I was wondering if there is a faster way to do the following.
>>>>
>>>>
>>>> I am minimizing a protein structure and one of the 'measurements" is
>>>> that
>>>> the minimized structure be as close to the starting value as possible.
>>>>
>>>>
>>>> Currently I use SVDSuperimposer.SVDSuperimposer() to calculate the RMSD
>>>> difference.
>>>>
>>>>
>>>> When I checked the various "energy terms" that are used to evaluate the
>>>> structure I find that the bottleneck is
>>>> indeed SVDSuperimposer.SVDSuperimposer().
>>>>
>>>>
>>>> Is there a way to do this more efficiently ?
>>>>
>>>>
>>>> Thank you
>>>> _______________________________________________
>>>> Biopython mailing list  -  Biopython at lists.open-bio.org
>>>> http://lists.open-bio.org/mailman/listinfo/biopython
>>>>
>>>
>>>
>>
>