Eugene Krissinel wrote:
> Thanks, I am reminded -- Eugene.
>
> On Tue, 28 Aug 2007, Paul Emsley wrote:
>
>> Dear Phil,
>>
>>
>> Phil Evans wrote:
>>> I note that Coot 0.3.2 still has a problem about removing cis peptides,
>>> in that "cis<->trans" does not remove the CISPEP lines from the output
>>> PDB file, so Refmac pops them back again. Is this on the list?
>>
>> It isn't. It should be. As far as I can see, this is mostly an mmdb
>> issue. (Again AFAICS) mmdb just writes out the CISPEP cards that it
>> reads. We need to have access to the CISPEP cards in an mmdb object
>> (typically to delete or add them as needed on writing) - I can't see how
>> to do that.
>>
>>
>>
>>> [in passing, cis peptide spurios can easily be introduced by Coot if the
>>> "add-planar-peptide-restraints" flag is off: shouldn't this be on by
>>> default?]
>> I agree. It will be default in 0.4 (and is default in current
>> pre-releases).
>>
Eugene,
I picked up your new code and integrated it. Thanks.
I can now test the coordinates against the header specification of CISPEPs.
To remove an incorrect CISPEP, I simply set it to NULL, do I? (which is
why you test for null in the example, I take it?).
Cheers,
Paul.
|