I have a postdoc here who is seeing curiously different results from the latest rmeasure (version 1.05) compared to a previous version (1.03).
This is a cryo structure where initial projection matching was done in SPIDER and then the structure was refined using FREALIGN.
rmeasure version 1.03 gives a resolution of 24 A.
rmeasure version 1.05 gives a resolution of 6 A, which would be nice but is certainly not true.
There is a new rmeasure input in 1.05 and not 1.03, the inner and outer radii automasking criteria. For the above result of 6 A, the value 0,0 was used as recommended.
If 0,140 is used instead (which are the same parameters that were used in FREALIGN), the resultant resolution is 28 A. The molecular mass that is calculated, however, is far larger than the actual structure.
I've been told that rmeasure isn't really reliable for structures of resolution greater than 15 A - is this still the case?
Yes, this is correct.
Yes, this is correct. Structures that have mainly low resolution signal will not be measured correctly. This is explained in the original publication. There might be a way around this problem but I have not had time to work on this. So the 24 or 28 A resolution might be close to the actual resolution but these values are probably not very reliable.