Jim's comment brings us back to the Dauter in us, or better the Dauter which should be in us. Data Data Data ! Before shooting take a moment and think about the experiment. How do you collect the best possible data from a bad crystal. Investing time at the beginning always pays off (unless the beam dumps before you completed your data set).
Jürgen - Jürgen Bosch Johns Hopkins Bloomberg School of Public Health Department of Biochemistry & Molecular Biology Johns Hopkins Malaria Research Institute 615 North Wolfe Street, W8708 Baltimore, MD 21205 Phone: +1-410-614-4742 Lab: +1-410-614-4894 Fax: +1-410-955-3655 http://web.mac.com/bosch_lab/<http://web.me.com/bosch_lab/> On Jan 28, 2011, at 11:05 AM, Jim Pflugrath wrote: You should know that your crystal mosaicity is a physical property of your crystals and the diffraction experiment. Generally, it is anisotropic though most programs output a single value. How can that single value describe what is really happening in your experimental data? You can do anything you want to with the processing programs. You can fix mosaicity to any value you want. You can restrict it to a small value to lie to the program that your spots are not overlapped. This should help completeness and redundancy while perhaps degrading accuracy. Will that help you solve the structure? Will that help to find the anomalous substructure? WIll that help to get an initial map for chain tracing? Will you get a better Rfree if you use data that is merged from several crystals? Will you get a better Rfree if you mix and match different mosaicities when processing the diffraction images from different crystals? These are all hypotheses that you can test. I am not sure how to test these hypotheses by querying the internet. ________________________________ From: CCP4 bulletin board [mailto:[email protected]] On Behalf Of Anastassis Perrakis Sent: Friday, January 28, 2011 8:11 AM To: [email protected]<mailto:[email protected]> Subject: Re: [ccp4bb] Merging data to increase multiplicity ... but, back to the main point, my advice would be to only limit the mosaicity, to get better completeness by avoiding overlaps. Its not ideal, in the sense that you would be over-estimating the partial fraction of most partial reflections, and thus systematically underestimating intensities. (I hope I got my overs and unders right here ...) But these errors would not matter much for refinement purposes, where you would rather have a slightly systematically wrong estimate for all data, rather than not have the 15% of the data at all. Or at least thats what I thought back in '99 refining MutS ... where I did refine a lot with both datasets and liked the 'fixed mosaicity' one better. A. On Jan 28, 2011, at 13:26, José Trincão wrote: Ah, yes, I was missing that. The statistics will be wrong. But in principle I will get an mtz with better data, because I am integrating more observations which would have been rejected by being missed at low resolution if the mosaicity was set too low or being rejected by overlaps at high resolution if the mosaicity is increased. So the question is - can I use this data for refinement? Or should I stick with the best of the datasets (the one with the highest completeness and multiplicity)? Thanks! Jose On Jan 28, 2011, at 28/1/11 - 11:59, Ian Tickle wrote: Jose - you're missing the fact that the same dataset processed in different ways are not statistically independent datasets! Increasing the multiplicity for independent data reduces the uncertainty because the calculation of the SU assumes statistical independence. Cheers -- Ian On Fri, Jan 28, 2011 at 11:46 AM, José Trincão <[email protected]<mailto:[email protected]>> wrote: Hello all, I have been trying to squeeze the most out of a bad data set (P1, anisotropic, crystals not reproducible). I had very incomplete data due to high mosaicity and lots of overlaps. The completeness was about 80% overall to ~3A. Yesterday I noticed that I could process the data much better fixing the mosaicity to 0.5 in imosflm. I got about 95% complete up to 2.5A but with a multiplicity of 1.7. I tried to integrate the same data fixing the mosaicity at different values ranging from 0.2 to 0.6 and saw the trend in completeness, Rmerge and multiplicity. Now, is there any reason why I should not just merge all these together and feed them to scala in order to increase multiplicity? Am I missing something? Thanks for any comments! Jose José Trincão, PhD CQFB@FCT-UNL 2829-516 Caparica, Portugal "It's very hard to make predictions... especially about the future" - Niels Bohr José Trincão, PhD CQFB@FCT-UNL 2829-516 Caparica, Portugal "It's very hard to make predictions... especially about the future" - Niels Bohr José Trincão, PhD CQFB@FCT-UNL 2829-516 Caparica, Portugal "It's very hard to make predictions... especially about the future" - Niels Bohr P please don't print this e-mail unless you really need to Anastassis (Tassos) Perrakis, Principal Investigator / Staff Member Department of Biochemistry (B8) Netherlands Cancer Institute, Dept. B8, 1066 CX Amsterdam, The Netherlands Tel: +31 20 512 1951 Fax: +31 20 512 1954 Mobile / SMS: +31 6 28 597791
