Re: [ccp4bb] Difficulty indexing diffraction, cell too small
Jason, How does the diffraction pattern look? Do you see multiple lattices or split spots? It seems to me that XDS isn't able to index it (correctly) possibly because of these things. It's always worth it to manually page through your images just to confirm. I've personally seen people just use XDS and assume everything went well, and run into similar problems. Kelly *** Kelly Daughtry, Ph.D. Post-Doctoral Fellow, Raetz Lab Biochemistry Department Duke University Alex H. Sands, Jr. Building 303 Research Drive RM 250 Durham, NC 27710 P: 919-684-5178 *** On Tue, May 17, 2011 at 5:43 PM, Jason Busby j.bu...@auckland.ac.nz wrote: Hi, I have a diffraction data-set from a hexagonal rod shaped crystal, to about 2.0 Å. The problem comes when I try to process the data - Mosflm won't index it, and XDS indexes it as P622, but the unit cell is too small to contain even a single molecule of my protein. I have tried integrating it in some different space groups that XDS suggested (P2, C2, P1) but in all cases the Rmerge and Rmeas are worse than for P622. If I scale in P622 (or any of the other space groups) I get odd results from the twinning tests. For example, the 4th moment of E (expected values of 2 for untwinned, 1.5 for perfect twin) is around 1.1, and the L-test and cumulative intensity distribution are unusual as well (uploaded images here: http://tinypic.com/r/65rfr7/7 and here: http://tinypic.com/r/30adgyr/7 ) Has anyone else had similar issues? Any ideas would be appreciated. Thanks, Jason. -- Jason Busby PhD Student Laboratory of Structural Biology School of Biological Sciences University of Auckland Thomas Building 110 3a Symonds St Private Bag 92019 Auckland 1142 New Zealand ph: +64 9 3737599 ext 83888 fx: +64 9 3737414
Re: [ccp4bb] Difficulty indexing diffraction, cell too small
There are a couple of tricks I know for dealing with datasets that are problematic to index (ignore if you've tried them): 1. double-check the machine parameters (detector to crystal distance and beam center in particular). 2. Index in mosflm using widely separated images (usually n,n+45,n+90) starting from a strong zone. 3. In really problematic cases, I've sometimes had to resort to manual spot picking (thankfully not for many years). Good luck, Pete Jason Busby wrote: Hi, I have a diffraction data-set from a hexagonal rod shaped crystal, to about 2.0 Å. The problem comes when I try to process the data - Mosflm won't index it, and XDS indexes it as P622, but the unit cell is too small to contain even a single molecule of my protein. I have tried integrating it in some different space groups that XDS suggested (P2, C2, P1) but in all cases the Rmerge and Rmeas are worse than for P622. If I scale in P622 (or any of the other space groups) I get odd results from the twinning tests. For example, the 4th moment of E (expected values of 2 for untwinned, 1.5 for perfect twin) is around 1.1, and the L-test and cumulative intensity distribution are unusual as well (uploaded images here: http://tinypic.com/r/65rfr7/7 and here: http://tinypic.com/r/30adgyr/7 ) Has anyone else had similar issues? Any ideas would be appreciated. Thanks, Jason. -- Jason Busby PhD Student Laboratory of Structural Biology School of Biological Sciences University of Auckland Thomas Building 110 3a Symonds St Private Bag 92019 Auckland 1142 New Zealand ph: +64 9 3737599 ext 83888 fx: +64 9 3737414
Re: [ccp4bb] Difficulty indexing diffraction, cell too small
Related to manual indexing, 20 years ago I wrote a program to transform a list of peaks from the original Xentronics detector to points in reciprocal space. The peaks were written as a PDB file so the peaks (now waters) could be displayed on any graphics program. It was very useful for seeing and separating multiple lattices. Is there an option in MOSFLM or D*Trek to output peaks in reciprocal x*, y*, z*? Doug Douglas H. Ohlendorf Phone: 612-624-8436 Professor FAX: 612-624-5121 Dept. of Biochemistry, Molecular Biology Biophysics Twin Cities Campus, University of Minnesota Lab web site: http://biosci.cbs.umn.edu/bmbb/ohlen_lab/index.html -Original Message- From: CCP4 bulletin board [mailto:CCP4BB@JISCMAIL.AC.UK] On Behalf Of Pete Meyer Sent: Wednesday, May 18, 2011 9:46 AM To: CCP4BB@JISCMAIL.AC.UK Subject: Re: [ccp4bb] Difficulty indexing diffraction, cell too small There are a couple of tricks I know for dealing with datasets that are problematic to index (ignore if you've tried them): 1. double-check the machine parameters (detector to crystal distance and beam center in particular). 2. Index in mosflm using widely separated images (usually n,n+45,n+90) starting from a strong zone. 3. In really problematic cases, I've sometimes had to resort to manual spot picking (thankfully not for many years). Good luck, Pete Jason Busby wrote: Hi, I have a diffraction data-set from a hexagonal rod shaped crystal, to about 2.0 Å. The problem comes when I try to process the data - Mosflm won't index it, and XDS indexes it as P622, but the unit cell is too small to contain even a single molecule of my protein. I have tried integrating it in some different space groups that XDS suggested (P2, C2, P1) but in all cases the Rmerge and Rmeas are worse than for P622. If I scale in P622 (or any of the other space groups) I get odd results from the twinning tests. For example, the 4th moment of E (expected values of 2 for untwinned, 1.5 for perfect twin) is around 1.1, and the L-test and cumulative intensity distribution are unusual as well (uploaded images here: http://tinypic.com/r/65rfr7/7 and here: http://tinypic.com/r/30adgyr/7 ) Has anyone else had similar issues? Any ideas would be appreciated. Thanks, Jason. -- Jason Busby PhD Student Laboratory of Structural Biology School of Biological Sciences University of Auckland Thomas Building 110 3a Symonds St Private Bag 92019 Auckland 1142 New Zealand ph: +64 9 3737599 ext 83888 fx: +64 9 3737414
Re: [ccp4bb] Difficulty indexing diffraction, cell too small
Yes, reciprocal lattice coordinates are available for reflections with d*TREK. My colleague has also written a reciprocal lattice viewer which takes image pixels and transforms them to a reciprocal lattice. I'm sure others have similar programs. But in this modern internet age, I would say enlist the help of experts with strange images. I think that usually saves one from going down an unproductive path. Jim -Original Message- From: CCP4 bulletin board [mailto:CCP4BB@JISCMAIL.AC.UK] On Behalf Of Doug Ohlendorf Sent: Wednesday, May 18, 2011 11:00 AM To: CCP4BB@JISCMAIL.AC.UK Subject: Re: [ccp4bb] Difficulty indexing diffraction, cell too small Related to manual indexing, 20 years ago I wrote a program to transform a list of peaks from the original Xentronics detector to points in reciprocal space. The peaks were written as a PDB file so the peaks (now waters) could be displayed on any graphics program. It was very useful for seeing and separating multiple lattices. Is there an option in MOSFLM or D*Trek to output peaks in reciprocal x*, y*, z*? Doug Douglas H. Ohlendorf Phone: 612-624-8436 Professor FAX: 612-624-5121 Dept. of Biochemistry, Molecular Biology Biophysics Twin Cities Campus, University of Minnesota Lab web site: http://biosci.cbs.umn.edu/bmbb/ohlen_lab/index.html -Original Message- From: CCP4 bulletin board [mailto:CCP4BB@JISCMAIL.AC.UK] On Behalf Of Pete Meyer Sent: Wednesday, May 18, 2011 9:46 AM To: CCP4BB@JISCMAIL.AC.UK Subject: Re: [ccp4bb] Difficulty indexing diffraction, cell too small There are a couple of tricks I know for dealing with datasets that are problematic to index (ignore if you've tried them): 1. double-check the machine parameters (detector to crystal distance and beam center in particular). 2. Index in mosflm using widely separated images (usually n,n+45,n+90) starting from a strong zone. 3. In really problematic cases, I've sometimes had to resort to manual spot picking (thankfully not for many years). Good luck, Pete Jason Busby wrote: Hi, I have a diffraction data-set from a hexagonal rod shaped crystal, to about 2.0 Å. The problem comes when I try to process the data - Mosflm won't index it, and XDS indexes it as P622, but the unit cell is too small to contain even a single molecule of my protein. I have tried integrating it in some different space groups that XDS suggested (P2, C2, P1) but in all cases the Rmerge and Rmeas are worse than for P622. If I scale in P622 (or any of the other space groups) I get odd results from the twinning tests. For example, the 4th moment of E (expected values of 2 for untwinned, 1.5 for perfect twin) is around 1.1, and the L-test and cumulative intensity distribution are unusual as well (uploaded images here: http://tinypic.com/r/65rfr7/7 and here: http://tinypic.com/r/30adgyr/7 ) Has anyone else had similar issues? Any ideas would be appreciated. Thanks, Jason. -- Jason Busby PhD Student Laboratory of Structural Biology School of Biological Sciences University of Auckland Thomas Building 110 3a Symonds St Private Bag 92019 Auckland 1142 New Zealand ph: +64 9 3737599 ext 83888 fx: +64 9 3737414
[ccp4bb] Difficulty indexing diffraction, cell too small
Hi, I have a diffraction data-set from a hexagonal rod shaped crystal, to about 2.0 Å. The problem comes when I try to process the data - Mosflm won't index it, and XDS indexes it as P622, but the unit cell is too small to contain even a single molecule of my protein. I have tried integrating it in some different space groups that XDS suggested (P2, C2, P1) but in all cases the Rmerge and Rmeas are worse than for P622. If I scale in P622 (or any of the other space groups) I get odd results from the twinning tests. For example, the 4th moment of E (expected values of 2 for untwinned, 1.5 for perfect twin) is around 1.1, and the L-test and cumulative intensity distribution are unusual as well (uploaded images here: http://tinypic.com/r/65rfr7/7 and here: http://tinypic.com/r/30adgyr/7 ) Has anyone else had similar issues? Any ideas would be appreciated. Thanks, Jason. -- Jason Busby PhD Student Laboratory of Structural Biology School of Biological Sciences University of Auckland Thomas Building 110 3a Symonds St Private Bag 92019 Auckland 1142 New Zealand ph: +64 9 3737599 ext 83888 fx: +64 9 3737414
Re: [ccp4bb] Difficulty indexing diffraction, cell too small
In general, the Rmerge and Rmeas should get better with a lower symmetry spacegroup, so that's weird. Maybe you didn't crystallize what you thought you crystallized. Do people runs gels anymore on crystals to get an idea of what's in the crystal or do they do MassSpec? I think another way to go at this is to make images available and have folks try to process your data for you. That's starting to become more common nowadays. Jim -Original Message- From: CCP4 bulletin board [mailto:CCP4BB@JISCMAIL.AC.UK] On Behalf Of Jason Busby Sent: Tuesday, May 17, 2011 4:44 PM To: CCP4BB@JISCMAIL.AC.UK Subject: [ccp4bb] Difficulty indexing diffraction, cell too small Hi, I have a diffraction data-set from a hexagonal rod shaped crystal, to about 2.0 Å. The problem comes when I try to process the data - Mosflm won't index it, and XDS indexes it as P622, but the unit cell is too small to contain even a single molecule of my protein. I have tried integrating it in some different space groups that XDS suggested (P2, C2, P1) but in all cases the Rmerge and Rmeas are worse than for P622. If I scale in P622 (or any of the other space groups) I get odd results from the twinning tests. For example, the 4th moment of E (expected values of 2 for untwinned, 1.5 for perfect twin) is around 1.1, and the L-test and cumulative intensity distribution are unusual as well (uploaded images here: http://tinypic.com/r/65rfr7/7 and here: http://tinypic.com/r/30adgyr/7 ) Has anyone else had similar issues? Any ideas would be appreciated. Thanks, Jason. -- Jason Busby PhD Student Laboratory of Structural Biology School of Biological Sciences University of Auckland Thomas Building 110 3a Symonds St Private Bag 92019 Auckland 1142 New Zealand ph: +64 9 3737599 ext 83888 fx: +64 9 3737414