Re: [ccp4bb] MR problem --molrep
Hi there, I think 55% completeness is insufficient - you really need to collect more data. Take a look at this movie from James Holton's website: http://ucxray.berkeley.edu/~jamesh/movies/osc.mpeg Stop the movie at 55% and check out how lousy the maps are compared to 95-100% - and I think these maps are generated with perfect phases, at 1.5A resolution. Molecular replacement programs and refinement programs are going to struggle with just over half a dataset - this is also why your R-factors are sky high. In addition, I know of cases where reviewers have cast doubt over structures (of a point mutant) with data of 80% completeness. I don't think a 55% complete dataset is going to get out there. I really think more data is the only way our of your predicament. I hope you have a few crystals ready to go! Good Luck, David. 2008/7/27 Carl Soja [EMAIL PROTECTED]: HI all, Thank you very much for your kindly answer. I also tried to use different program like phaser, amore, and some MR servers. actualy, I don`t know how many molecule in AU clearly. 2-6 molecules in 76%- 35% of solvent content. 4mer is 50% of solvent content. the homolgous structure has high solvent content over 70%. when i used phaser program to do MR, I can get several solutions but not good structure conformation. I don`t know how to use phaser do automatic more than 2 molecules search like molrep program. As spacegroup, I checked the C222 and C2221. the phaser program can give both solutions when I changed the search model or maximum solution limit. my question is how i can search more than 2 molecules by phaser program like molrp only input over 2. by the way, my diffraction data have low completeness(55%) and low redandancy(1.8). Thanks in advance!! carl roja -- David C. Briggs PhD Father Crystallographer http://www.dbriggs.talktalk.net AIM ID: dbassophile
Re: [ccp4bb] MR problem --molrep
HI all, Thank you very much for your kindly answer. I also tried to use different program like phaser, amore, and some MR servers. actualy, I don`t know how many molecule in AU clearly. 2-6 molecules in 76%- 35% of solvent content. 4mer is 50% of solvent content. the homolgous structure has high solvent content over 70%. when i used phaser program to do MR, I can get several solutions but not good structure conformation. I don`t know how to use phaser do automatic more than 2 molecules search like molrep program. As spacegroup, I checked the C222 and C2221. the phaser program can give both solutions when I changed the search model or maximum solution limit. my question is how i can search more than 2 molecules by phaser program like molrp only input over 2. by the way, my diffraction data have low completeness(55%) and low redandancy(1.8). Thanks in advance!! carl roja
[ccp4bb] MR problem --molrep
Dear all I tried to solve one structure by ccp4i molrep(resolution at 3.0 A, space group C222, sequence ID 30%). I can get a good Rfactor 0.528 at first translation function. However, the second translation function Rfac is 0.526, the third is 0.525, the fourth is 0.525. All of the translation function Rfacs are too closed. I changed the model and minimum resolution for search, the Rfactor closed no any improved. My structure estmates four molecules in the aymmetric unit. This is my first time found the closed Rfac by molrep. From the low translation function Rfac, it seems that it is correct solution. I checked the solution and found some clashes in the structure.I am not sure why the Rfactor too closed in next the molecule search? I know this is unusual solution by molrep. Does it mean the diffraction data has problem? Any suggestions are welcome.Thank you in advance! Carl soja
Re: [ccp4bb] MR problem --molrep
Hi, I dare to say about the possible way to do molrep from my recent experience. You can choose rotation and translation function job at first and do the self rotation fuction (for multimer) after that.Each of these run will generate two different outputs *_rf.molrep_rf ans *_srf.molrep_rf. Make the third run inputting both of these peak values. Give NMON as 4. You can also use the protein sequence hare. Then refine it with Refmac, that again should decrease your R factor. I will be highly benefitted if some one point out my mistake. regards Debajyoti On Fri, 25 Jul 2008 Carl Soja wrote : Dear all I tried to solve one structure by ccp4i molrep(resolution at 3.0 A, space group C222, sequence ID 30%). I can get a good Rfactor 0.528 at first translation function. However, the second translation function Rfac is 0.526, the third is 0.525, the fourth is 0.525. All of the translation function Rfacs are too closed. I changed the model and minimum resolution for search, the Rfactor closed no any improved. My structure estmates four molecules in the aymmetric unit. This is my first time found the closed Rfac by molrep. From the low translation function Rfac, it seems that it is correct solution. I checked the solution and found some clashes in the structure.I am not sure why the Rfactor too closed in next the molecule search? I know this is unusual solution by molrep. Does it mean the diffraction data has problem? Any suggestions are welcome.Thank you in advance! Carl soja
Re: [ccp4bb] MR problem --molrep
Dear Debajyoti Thank you very much for your help. I input the each peak values as 10 and carried out the rotation and translation function again. I didn't get a improved solution by molrep. can i edit the self-rotaion function and translation function solution file as input ? best regards, carl soja Hi, I dare to say about the possible way to do molrep from my recent experience. You can choose rotation and translation function job at first and do the self rotation fuction (for multimer) after that.Each of these run will generate two different outputs *_rf.molrep_rf ans *_srf.molrep_rf. Make the third run inputting both of these peak values. Give NMON as 4. You can also use the protein sequence hare. Then refine it with Refmac, that again should decrease your R factor. I will be highly benefitted if some one point out my mistake. regards Debajyoti On Fri, 25 Jul 2008 Carl Soja wrote : Dear all I tried to solve one structure by ccp4i molrep(resolution at 3.0 A, space group C222, sequence ID 30%). I can get a good Rfactor 0.528 at first translation function. However, the second translation function Rfac is 0.526, the third is 0.525, the fourth is 0.525. All of the translation function Rfacs are too closed. I changed the model and minimum resolution for search, the Rfactor closed no any improved. My structure estmates four molecules in the aymmetric unit. This is my first time found the closed Rfac by molrep. From the low translation function Rfac, it seems that it is correct solution. I checked the solution and found some clashes in the structure.I am not sure why the Rfactor too closed in next the molecule search? I know this is unusual solution by molrep. Does it mean the diffraction data has problem? Any suggestions are welcome.Thank you in advance! Carl soja
Re: [ccp4bb] MR problem --molrep
Carl Soja wrote: Dear all I tried to solve one structure by ccp4i molrep(resolution at 3.0 A, space group C222, sequence ID 30%). I can get a good Rfactor 0.528 at first translation function. However, the second translation function Rfac is 0.526, the third is 0.525, the fourth is 0.525. All of the translation function Rfacs are too closed. I changed the model and minimum resolution for search, the Rfactor closed no any improved. My structure estmates four molecules in the aymmetric unit. This is my first time found the closed Rfac by molrep. From the low translation function Rfac, it seems that it is correct solution. I checked the solution and found some clashes in the structure.I am not sure why the Rfactor too closed in next the molecule search? I know this is unusual solution by molrep. Does it mean the diffraction data has problem? Any suggestions are welcome.Thank you in advance! Carl soja To be honest, I would try both Phaser and EPMR (now Open-EPMR) first to do MR. Both especially excel at finding good MR solutions for multimers, and are very fast as well. These programs can find solutions that are very difficult to find using other rotation-translation programs, including MOLREP. In my experience (using Phaser and EPMR) reasonable MR solutions almost always have an R-factor under 50%. Cheers, -- Roger S. Rowlett Professor Colgate University Presidential Scholar Department of Chemistry Colgate University 13 Oak Drive Hamilton, NY 13346 tel: (315)-228-7245 ofc: (315)-228-7395 fax: (315)-228-7935 email: [EMAIL PROTECTED]
Re: [ccp4bb] MR problem --molrep
Perhaps obvious - are you sure the space group is C222 not C2221? Phil On 25 Jul 2008, at 14:19, Roger Rowlett wrote: Carl Soja wrote: Dear all I tried to solve one structure by ccp4i molrep(resolution at 3.0 A, space group C222, sequence ID 30%). I can get a good Rfactor 0.528 at first translation function. However, the second translation function Rfac is 0.526, the third is 0.525, the fourth is 0.525. All of the translation function Rfacs are too closed. I changed the model and minimum resolution for search, the Rfactor closed no any improved. My structure estmates four molecules in the aymmetric unit. This is my first time found the closed Rfac by molrep. From the low translation function Rfac, it seems that it is correct solution. I checked the solution and found some clashes in the structure.I am not sure why the Rfactor too closed in next the molecule search? I know this is unusual solution by molrep. Does it mean the diffraction data has problem? Any suggestions are welcome.Thank you in advance! Carl soja To be honest, I would try both Phaser and EPMR (now Open-EPMR) first to do MR. Both especially excel at finding good MR solutions for multimers, and are very fast as well. These programs can find solutions that are very difficult to find using other rotation- translation programs, including MOLREP. In my experience (using Phaser and EPMR) reasonable MR solutions almost always have an R- factor under 50%. Cheers, -- Roger S. Rowlett Professor Colgate University Presidential Scholar Department of Chemistry Colgate University 13 Oak Drive Hamilton, NY 13346 tel: (315)-228-7245 ofc: (315)-228-7395 fax: (315)-228-7935 email: [EMAIL PROTECTED]
Re: [ccp4bb] MR problem --molrep
In addition to Bert's statistics argument (I've never had the pleasure of working w/ a C222 crystal), do check your self-Patterson map. I recently had a difficult MR case; the crystal masqueraded as P21212 or P2221, but it was actually P212121 with the two molecules in the AU related by a non-crystallographic translation. Dave Borhani -Original Message- From: CCP4 bulletin board [mailto:[EMAIL PROTECTED] On Behalf Of Roger Rowlett Sent: Friday, July 25, 2008 10:30 AM To: CCP4BB@JISCMAIL.AC.UK Subject: Re: [ccp4bb] MR problem --molrep Good point. One should always check all the screw axis combos. Not fun for P422, but Phaser makes this very easy, as it will do all the possible screw axis combinations in one job. The correct space group and MR solution is usually very obvious if the model is acceptable and the MR parameters have been set reasonably. I think Open-EPMR will also examine screw axis combinations now, but I've never used it to do this. Cheers, Roger Rowlett Phil Evans wrote: Perhaps obvious - are you sure the space group is C222 not C2221? Phil On 25 Jul 2008, at 14:19, Roger Rowlett wrote: Carl Soja wrote: Dear all I tried to solve one structure by ccp4i molrep(resolution at 3.0 A, space group C222, sequence ID 30%). I can get a good Rfactor 0.528 at first translation function. However, the second translation function Rfac is 0.526, the third is 0.525, the fourth is 0.525. All of the translation function Rfacs are too closed. I changed the model and minimum resolution for search, the Rfactor closed no any improved. My structure estmates four molecules in the aymmetric unit. This is my first time found the closed Rfac by molrep. From the low translation function Rfac, it seems that it is correct solution. I checked the solution and found some clashes in the structure.I am not sure why the Rfactor too closed in next the molecule search? I know this is unusual solution by molrep. Does it mean the diffraction data has problem? Any suggestions are welcome.Thank you in advance! Carl soja To be honest, I would try both Phaser and EPMR (now Open-EPMR) first to do MR. Both especially excel at finding good MR solutions for multimers, and are very fast as well. These programs can find solutions that are very difficult to find using other rotation- translation programs, including MOLREP. In my experience (using Phaser and EPMR) reasonable MR solutions almost always have an R- factor under 50%. Cheers, -- -- -- Roger S. Rowlett Professor Colgate University Presidential Scholar Department of Chemistry Colgate University 13 Oak Drive Hamilton, NY 13346 tel: (315)-228-7245 ofc: (315)-228-7395 fax: (315)-228-7935 email: [EMAIL PROTECTED]
