Hi Paul, thanks for your quick response.
On 11 March 2014 08:38, Paul Emsley <[email protected]> wrote: > On 11/03/14 07:14, Michael Engel wrote: > >> I can't reproduce the results I get by using find waters from within coot >> when running findwaters on the command line on the same pdb/mtz. I have >> double checked all the command line switches are the same and compared the >> output in the log files. Nothing obviously different but I am getting >> nearly 3x as many waters when running findwaters externally compared to >> starting it from the GUI, so as if the sigma was on a different scale. >> > > So, do the first 1/3 of the waters from findwaters match those from the > GUI? Some do some don't. > Does coot and findwaters report the same "density cut-off"? > Here are the relevant lines: findwaters: INFO:: find_clusters map_rms is 0.125677 INFO:: Using density cut-off: 0.377032 (3 sigma) (mean -0.00607296 stdev: 0.0929294) INFO:: Using water to protein distance limits: 2.4 3.2 INFO:: found 150 waters in water fitting coot find waters: INFO:: find_clusters map_rms is 0.128606 INFO:: Using density cut-off: 0.385819 (3 sigma) (mean -0.0055491 stdev: 0.0881163) INFO:: Using water to protein distance limits: 2.4 3.2 INFO:: found 56 waters in water fitting > And the other 2/3 are in smaller map peaks? > > Yep. Michael
