*** For details on how to be removed from this list visit the *** *** CCP4 home page http://www.ccp4.ac.uk ***
You could try d*TREK.
Definitely worth a try. I once had a case where the best I could get with HKL2000 was about 65% due to overlaps in the highly mosaic dataset. d*TREK, after some fiddling
(and invaluable input from Jim Pflugrath) eventually gave 98% completeness with decent I/sigma.
I tried using the elliptical spot feature of HKL2000 but this doesn't seem to be ideal as the spots are not always elongated in the same direction.
You don't have to stick to the same spot shape in your entire dataset. You can break it into several separate ones, integrate them using different ellipsoid spots (when done carefully, this should reduce the number of overlaps considerably) and then scale all .x files together.
Dima
