-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Dear Antony,
I agree, modelling entirely separate entities with one occupancy for all atoms inside make chemically much more sense and it has the advantage that the refinement program can move e.g. ADP to a slightly different position than ATP - it is unlikely that the common atoms between the two molecules are exactly in the same position. Best, Tim On 05/25/2013 12:34 PM, Antony Oliver wrote: > Kavya, > > Presumably your enzyme is turning over the ATP? Driving it towards > ADP? > > In that case I suspect that you may have a mixture of ATP and ADP > (and possibly contaminating AMP). You could either model both ATP > and ADP, and set relative occupancies of both (add up to 1). Or, > you could assign different occupancies to the phosphate groups > alone. I personally think the former is probably the correct > method? Comments CCP4 community? > > Many regards, Antony. > > > On 25 May 2013, at 11:29, <[email protected]> wrote: > >> Dear Sir, >> >> That is true the ligand is ATP, the occupancy problem comes only >> for the phosphates. This ATP tends to get cleaved to ADP and AMP >> in other complexes I got. So in this case do you suggest keeping >> different occupancies for nucleoside and phosphate groups? I am >> not aware of any publication with this scenario so I am not sure >> whether it is right. >> >> >> Thank you Regards Kavya >> >>> Dear Kavyashree, >>> >>> It is possible that your bound ligand (for which you have >>> strong electron density) is actually a break-down of the parent >>> compound? We have seen this a couple of times now. >>> >>> Also - are the poorly refining areas (those with negative >>> density) part of a pendant ring connected by a conformationally >>> unrestricted bond? These quite often have poor density. >>> >>> Hard to judge without seeing the actual density - but >>> understand why! >>> >>> Regards, Antony. >>> >>> >>> Sent from my iPhone >>> >>> On 25 May 2013, at 10:40, "Kavyashree Manjunath" >>> <[email protected]> wrote: >>> >>>> Dear Sir, >>>> >>>> Thank you all for your kind advice and clarifications. I will >>>> keep the occupancy 1.0 for both the ligands and refine >>>> without considering the negative density in this case. >>>> >>>> Thanking you Regards Kavya >>>> >>>> >>>> > Dear Kavya, > > I don't see much sense in having different occupancies within the > same molecule (unless one atoms sits on a special position, but > then refmac will take care of it). > > If positive density comes up it's a good sign the ligand really is > there. At 2.2A I would not be too surprised some atoms show less > density than others (but don't look too much at the map with a > sigma level < 1: you are going to see what you want to see). > > It's difficult to judge without sitting next to you, so my advice > is try to model it as good as your knowledge permits, and do take > your chemical knowledge into account when doing so! > > Best, Tim > > On 05/25/2013 06:55 AM, Kavyashree Manjunath wrote: >>>>>>> Dear users, >>>>>>> >>>>>>> I tried giving occupancy of 0.65 and 0.6 respectively >>>>>>> for all atoms of the two ligands and refined. Now after >>>>>>> refinement, the atoms of ligand does not have negative >>>>>>> density but those which did not have negative density >>>>>>> previously appear positive. So what do I need to do? >>>>>>> under what circumstances does a ligand have different >>>>>>> occupancies for different atoms or for a group of >>>>>>> atoms. Any such references are very much welcome. >>>>>>> >>>>>>> Thank you Regards Kavya >>>>>>> >>>>>>>> Sir, >>>>>>>> >>>>>>>> Yes it is around ligand. The average B-factor of one >>>>>>>> of the ligand is 36.78, of which one of the atom has >>>>>>>> occupancy B factor 0.58 39.37 0.56 >>>>>>>> 38.77 0.87 37.00 Three atoms are of same type. The >>>>>>>> other ligand's overall Bfactor is 17.64. occupancy >>>>>>>> B factor 1.00 19.29 0.64 23.90 >>>>>>>> Two atoms are of same type. >>>>>>>> >>>>>>>> So in the present case should I put the occupancy of >>>>>>>> 0.56 for all atoms of ligand-1 and 0.64 for all atoms >>>>>>>> of ligand-2 and refine? >>>>>>>> >>>>>>>> I mean will it be wrong to put different occupancies >>>>>>>> for different atoms of same ligand? >>>>>>>> >>>>>>>> I could not see any alternate densities coming up for >>>>>>>> those atoms which did not have densities. But 2FoFc >>>>>>>> map would appear around these atoms at 0.7 sigma at >>>>>>>> the same position where the atoms are present. >>>>>>>> >>>>>>>> Thank you Regards Kavya >>>>>>>> >>>>>>>>> Hi Kavya, >>>>>>>>> >>>>>>>>> If I understand you correctly (from title and >>>>>>>>> text), you meant your negative FoFc was around your >>>>>>>>> ligand, is that right? I wonder if this is a case >>>>>>>>> in which the ligand has an occupancy below 1, but >>>>>>>>> was modeled as 1, so the refinement program had to >>>>>>>>> give it a high B factor to compensate that, which >>>>>>>>> results in the electron density bleeding into >>>>>>>>> nearby space where there should not be so much of >>>>>>>>> it. If you want to test this hypothesis, one thing >>>>>>>>> you can try is to set the occupancy to 0.25, >>>>>>>>> 0.5,0.75 and so on, and refine a few cycles to see >>>>>>>>> what happens to the maps. Also, what's the B factor >>>>>>>>> of the ligand, and what's the B of the nearby >>>>>>>>> protein atoms? The difference between them can also >>>>>>>>> give you some hint for guessing the ligand's >>>>>>>>> occupancy. Some refinement programs(phenix.refine >>>>>>>>> and shelx) can also let you refine the ligand's >>>>>>>>> occupancy. >>>>>>>>> >>>>>>>>> As to the "missing" atoms, that could be caused by >>>>>>>>> alternative conformations of the ligand - assuming >>>>>>>>> you have already done a thorough refinement. >>>>>>>>> >>>>>>>>> Zhijie >>>>>>>>> >>>>>>>>> -------------------------------------------------- >>>>>>>>> From: "Kavyashree Manjunath" >>>>>>>>> <[email protected]> Sent: Friday, May 24, 2013 >>>>>>>>> 12:50 PM To: <[email protected]> Subject: >>>>>>>>> [ccp4bb] Negative FoFc around ligand >>>>>>>>> >>>>>>>>>> Dear users, >>>>>>>>>> >>>>>>>>>> I am using refmac 5.7.0029 for refining a >>>>>>>>>> structure (resolution 2.2 Ang) bound to 2 >>>>>>>>>> ligands. After MR There is a very clear density >>>>>>>>>> of ligands but after refinement, I get negative >>>>>>>>>> fofc map near one of the ligand upto 5 sigma. >>>>>>>>>> However its 2fofc map covers the whole ligand. >>>>>>>>>> Also for the other ligand, I do not see any 2fofc >>>>>>>>>> density (at 3 sigma) for 2 atoms, without these >>>>>>>>>> atoms the ligand is unrealistic. But the density >>>>>>>>>> comes up around these at around 0.7 sigma. >>>>>>>>>> Overall completeness is 99.9% Rmerge 7.5% >>>>>>>>>> >>>>>>>>>> What else I need to check in the data. Kindly >>>>>>>>>> provide some suggestions. >>>>>>>>>> >>>>>>>>>> Thanking you Regards Kavya >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> -- This message has been scanned for viruses and >>>>>>>>>> dangerous content by MailScanner, and is believed >>>>>>>>>> to be clean. >>>>>>>>> >>>>>>>>> >>>>>>>>> -- This message has been scanned for viruses and >>>>>>>>> dangerous content by MailScanner, and is believed >>>>>>>>> to be clean. >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> -- This message has been scanned for viruses and >>>>>>>> dangerous content by MailScanner, and is believed to >>>>>>>> be clean. > >>>>> >>>>> -- This message has been scanned for viruses and dangerous >>>>> content by MailScanner, and is believed to be clean. >>>> >>>> >>>> >>>> -- This message has been scanned for viruses and dangerous >>>> content by MailScanner, and is believed to be clean. >>> >>> -- This message has been scanned for viruses and dangerous >>> content by MailScanner, and is believed to be clean. >>> >>> >> >> >> >> -- This message has been scanned for viruses and dangerous >> content by MailScanner, and is believed to be clean. >> > - -- Dr Tim Gruene Institut fuer anorganische Chemie Tammannstr. 4 D-37077 Goettingen GPG Key ID = A46BEE1A -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.12 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iD8DBQFRoKi6UxlJ7aRr7hoRApC5AJ4xpCWCzM7rDtZYeHNJNBvh8lsgdQCeIxc7 14cm2XkebzvgDH0BsTr4M+o= =fdnh -----END PGP SIGNATURE-----
