Re: [ccp4bb] PST in refinement
One question - do you have two pairs of molecules, each related by the PST or is there extra non-crystallographic translation ? Averaging or NCS restraints dont give much extra information for molecules in the same orientation. Certainly any pseudo translation will generate sets of weak and strong reflections but unless you have a simple fraction in each direction it is not easy to try to select a sub-set for refinement. Do all the data sets merge together reasonably? You probably just have to slog it through - do you have any experimental phasing at all? Eleanor Maruf Ali wrote: Dear all I have recently collected several datasets on different crystals of a particular protein with a resolution range form 2.4 - 3.2A. All datasets seem to process well in p21 with a unit cell of 109.6 83.1 115.87 90 94.8 90, and this space group is further supported by analysis with the program pointless. The dataset have very reasonable statistics and Rmerge values, with no indication of twinning. Analysis of the self patterson indicated a 43% off origin peak at 0.3 0.5 0.47. This was further flagged by pointless and molrep as a Psuedo cell translation (PST). Looking at the systematic absences there are some unusually strong and weak peaks. Initially after some toiling with molecular replacement, there was a clear solution with four molecules in the asymmetric unit. The maps generated were good enough to build the core of the protein but do not look like maps generated from data at 2.4A-3.2. Further more the free R is stuck at around 40%, and there is no difference in free R when I apply ncs or not or any difference in the maps. After building by hand and with phenix autobuild there is still no difference in maps and Rfree. I have read papers where labs have successfully refined PST data by separating the reflections according to the PST. Oksanen et al 2006 acta D62 1369-1374: Poy et al 2001 NSMB Vol 8 no12 pg 1053: VajDos et al protein science 1997 6;2297 My specific question are Firstly how would I deal with refining PST data? (assuming this is the problem). Second with off origin peak of 0.3 0.5 0.47 how would I separate the reflections? Thirdly any comments would be valued Thank you in advance Maruf Dr Maruf Ali Section of Structural Biology, Institute of Cancer Research, 237 Fulham road, London. SW3 6JB The Institute of Cancer Research: Royal Cancer Hospital, a charitable Company Limited by Guarantee, Registered in England under Company No. 534147 with its Registered Office at 123 Old Brompton Road, London SW7 3RP. This e-mail message is confidential and for use by the addressee only. If the message is received by anyone other than the addressee, please return the message to the sender by replying to it and then delete the message from your computer and network.
Re: [ccp4bb] PST in refinement
I have four molecules arranged such that molecules a and b are related by PST and adopt the same orientation as do molecules c and d, but there is no pst between the two pairs. All molecules lie in the same plane. A B top view C D from side view / X \ I hope that the description is clear. Any further comments would be appreciated. Maruf On 15 Jul 2008, at 12:28, Eleanor Dodson wrote: One question - do you have two pairs of molecules, each related by the PST or is there extra non-crystallographic translation ? Averaging or NCS restraints dont give much extra information for molecules in the same orientation. Certainly any pseudo translation will generate sets of weak and strong reflections but unless you have a simple fraction in each direction it is not easy to try to select a sub-set for refinement. Do all the data sets merge together reasonably? You probably just have to slog it through - do you have any experimental phasing at all? Eleanor Maruf Ali wrote: Dear all I have recently collected several datasets on different crystals of a particular protein with a resolution range form 2.4 - 3.2A. All datasets seem to process well in p21 with a unit cell of 109.6 83.1 115.87 90 94.8 90, and this space group is further supported by analysis with the program pointless. The dataset have very reasonable statistics and Rmerge values, with no indication of twinning. Analysis of the self patterson indicated a 43% off origin peak at 0.3 0.5 0.47. This was further flagged by pointless and molrep as a Psuedo cell translation (PST). Looking at the systematic absences there are some unusually strong and weak peaks. Initially after some toiling with molecular replacement, there was a clear solution with four molecules in the asymmetric unit. The maps generated were good enough to build the core of the protein but do not look like maps generated from data at 2.4A-3.2. Further more the free R is stuck at around 40%, and there is no difference in free R when I apply ncs or not or any difference in the maps. After building by hand and with phenix autobuild there is still no difference in maps and Rfree. I have read papers where labs have successfully refined PST data by separating the reflections according to the PST. Oksanen et al 2006 acta D62 1369-1374: Poy et al 2001 NSMB Vol 8 no12 pg 1053: VajDos et al protein science 1997 6;2297 My specific question are Firstly how would I deal with refining PST data? (assuming this is the problem). Second with off origin peak of 0.3 0.5 0.47 how would I separate the reflections? Thirdly any comments would be valued Thank you in advance Maruf Dr Maruf Ali Section of Structural Biology, Institute of Cancer Research, 237 Fulham road, London. SW3 6JB The Institute of Cancer Research: Royal Cancer Hospital, a charitable Company Limited by Guarantee, Registered in England under Company No. 534147 with its Registered Office at 123 Old Brompton Road, London SW7 3RP. This e-mail message is confidential and for use by the addressee only. If the message is received by anyone other than the addressee, please return the message to the sender by replying to it and then delete the message from your computer and network. Dr Maruf Ali Section of Structural Biology, Institute of Cancer Research, 237 Fulham road, London. SW3 6JB The Institute of Cancer Research: Royal Cancer Hospital, a charitable Company Limited by Guarantee, Registered in England under Company No. 534147 with its Registered Office at 123 Old Brompton Road, London SW7 3RP. This e-mail message is confidential and for use by the addressee only. If the message is received by anyone other than the addressee, please return the message to the sender by replying to it and then delete the message from your computer and network.
Re: [ccp4bb] PST in refinement
Hi Maruf, 1) Why the low-resolution data was collected to 3.2 Å? 2) Is it possible the real space group is P2 with a NCS which making it look like P21, especially when the PST has a value of exact 0.5. Lijun On Jul 14, 2008, at 5:42 AM, Maruf Ali wrote: Dear all I have recently collected several datasets on different crystals of a particular protein with a resolution range form 2.4 - 3.2A. All datasets seem to process well in p21 with a unit cell of 109.6 83.1 115.87 90 94.8 90, and this space group is further supported by analysis with the program pointless. The dataset have very reasonable statistics and Rmerge values, with no indication of twinning. Analysis of the self patterson indicated a 43% off origin peak at 0.3 0.5 0.47. This was further flagged by pointless and molrep as a Psuedo cell translation (PST). Looking at the systematic absences there are some unusually strong and weak peaks. Initially after some toiling with molecular replacement, there was a clear solution with four molecules in the asymmetric unit. The maps generated were good enough to build the core of the protein but do not look like maps generated from data at 2.4A-3.2. Further more the free R is stuck at around 40%, and there is no difference in free R when I apply ncs or not or any difference in the maps. After building by hand and with phenix autobuild there is still no difference in maps and Rfree. I have read papers where labs have successfully refined PST data by separating the reflections according to the PST. Oksanen et al 2006 acta D62 1369-1374: Poy et al 2001 NSMB Vol 8 no12 pg 1053: VajDos et al protein science 1997 6;2297 My specific question are Firstly how would I deal with refining PST data? (assuming this is the problem). Second with off origin peak of 0.3 0.5 0.47 how would I separate the reflections? Thirdly any comments would be valued Thank you in advance Maruf Dr Maruf Ali Section of Structural Biology, Institute of Cancer Research, 237 Fulham road, London. SW3 6JB The Institute of Cancer Research: Royal Cancer Hospital, a charitable Company Limited by Guarantee, Registered in England under Company No. 534147 with its Registered Office at 123 Old Brompton Road, London SW7 3RP. This e-mail message is confidential and for use by the addressee only. If the message is received by anyone other than the addressee, please return the message to the sender by replying to it and then delete the message from your computer and network. Lijun Liu, PhD Institute of Molecular Biology HHMI Department of Physics University of Oregon Eugene, OR 97403 541-346-4080
[ccp4bb] PST in refinement
Dear all I have recently collected several datasets on different crystals of a particular protein with a resolution range form 2.4 - 3.2A. All datasets seem to process well in p21 with a unit cell of 109.6 83.1 115.87 90 94.8 90, and this space group is further supported by analysis with the program pointless. The dataset have very reasonable statistics and Rmerge values, with no indication of twinning. Analysis of the self patterson indicated a 43% off origin peak at 0.3 0.5 0.47. This was further flagged by pointless and molrep as a Psuedo cell translation (PST). Looking at the systematic absences there are some unusually strong and weak peaks. Initially after some toiling with molecular replacement, there was a clear solution with four molecules in the asymmetric unit. The maps generated were good enough to build the core of the protein but do not look like maps generated from data at 2.4A-3.2. Further more the free R is stuck at around 40%, and there is no difference in free R when I apply ncs or not or any difference in the maps. After building by hand and with phenix autobuild there is still no difference in maps and Rfree. I have read papers where labs have successfully refined PST data by separating the reflections according to the PST. Oksanen et al 2006 acta D62 1369-1374: Poy et al 2001 NSMB Vol 8 no12 pg 1053: VajDos et al protein science 1997 6;2297 My specific question are Firstly how would I deal with refining PST data? (assuming this is the problem). Second with off origin peak of 0.3 0.5 0.47 how would I separate the reflections? Thirdly any comments would be valued Thank you in advance Maruf Dr Maruf Ali Section of Structural Biology, Institute of Cancer Research, 237 Fulham road, London. SW3 6JB The Institute of Cancer Research: Royal Cancer Hospital, a charitable Company Limited by Guarantee, Registered in England under Company No. 534147 with its Registered Office at 123 Old Brompton Road, London SW7 3RP. This e-mail message is confidential and for use by the addressee only. If the message is received by anyone other than the addressee, please return the message to the sender by replying to it and then delete the message from your computer and network.
Re: [ccp4bb] PST in refinement
Dear Maruf, I had a similar case, also to 2.4A. Those translations are often caused by a non-crystallographic 2-fold rotation which is parallel to the crystallographic 2-fold (there is an example in ruppweb as to why that creates the translation and how to interpret it: http:// www.ruppweb.org/Xray/101index.html under NCS with native Patterson maps). In your particular case, though, have you considered wether your absences match the strong-weak pattern created by your non- crystallographic translation? In that case maybe you don't have a 21 axis and your space group is P2 or C2. If your space group is right, the non crystallographic rotation will dominate your molecular replacement and your refinement and make it difficult to spot good solutions from bad ones, since as long as MR solutions recreate the non-crystallographic rotation (the axis, not necessarily the correct position of your molecules around that rotation axis), they will recreate the translation peak. I even got a series of totally different MR solutions, all of them with beautiful packing! For me it turned out that I had the correct MR solution for months, but the refinement was stalling. Running refmac for 200 cycles (not less!) with very very tight geometric restraints restraints produced a distribution of B-factors that pointed at which regions were right (low Bs) and which wrong (high Bs). Removal of the regions with high Bs followed by further refinement immediately produced good maps, showing the new regions, and refinement proceeded normally from there. Later on I could move back to normal weights, too. Only after cropping the bad regions will the difference maps really show anything (assuming the rest is right, of course). Just based on one case, but your case is s similar to mine! Good luck, Jose Antonio Cuesta-Seijo. ** Jose Antonio Cuesta-Seijo Cancer Genomics and Proteomics Ontario Cancer Institute, UHN MaRS TMDT Room 4-902M 101 College Street M5G 1L7 Toronto, ON, Canada Phone: (416)581-7544 Fax: (416)581-7562 email: [EMAIL PROTECTED] ** On Jul 14, 2008, at 8:42 AM, Maruf Ali wrote: Dear all I have recently collected several datasets on different crystals of a particular protein with a resolution range form 2.4 - 3.2A. All datasets seem to process well in p21 with a unit cell of 109.6 83.1 115.87 90 94.8 90, and this space group is further supported by analysis with the program pointless. The dataset have very reasonable statistics and Rmerge values, with no indication of twinning. Analysis of the self patterson indicated a 43% off origin peak at 0.3 0.5 0.47. This was further flagged by pointless and molrep as a Psuedo cell translation (PST). Looking at the systematic absences there are some unusually strong and weak peaks. Initially after some toiling with molecular replacement, there was a clear solution with four molecules in the asymmetric unit. The maps generated were good enough to build the core of the protein but do not look like maps generated from data at 2.4A-3.2. Further more the free R is stuck at around 40%, and there is no difference in free R when I apply ncs or not or any difference in the maps. After building by hand and with phenix autobuild there is still no difference in maps and Rfree. I have read papers where labs have successfully refined PST data by separating the reflections according to the PST. Oksanen et al 2006 acta D62 1369-1374: Poy et al 2001 NSMB Vol 8 no12 pg 1053: VajDos et al protein science 1997 6;2297 My specific question are Firstly how would I deal with refining PST data? (assuming this is the problem). Second with off origin peak of 0.3 0.5 0.47 how would I separate the reflections? Thirdly any comments would be valued Thank you in advance Maruf Dr Maruf Ali Section of Structural Biology, Institute of Cancer Research, 237 Fulham road, London. SW3 6JB The Institute of Cancer Research: Royal Cancer Hospital, a charitable Company Limited by Guarantee, Registered in England under Company No. 534147 with its Registered Office at 123 Old Brompton Road, London SW7 3RP. This e-mail message is confidential and for use by the addressee only. If the message is received by anyone other than the addressee, please return the message to the sender by replying to it and then delete the message from your computer and network.