Hey Pavel, Many thanks for your reply. The inputs are the same i.e. same model and same map were used as inputs for either Phenix version. Chancie.

Am 19.11.2019 um 00:23 schrieb Pavel Afonine :

Hi Chancievan,

comparing the two files you attached I see you are looking at two different model and map files. Is this the case?

If you are comparing two different Phenix versions at least you need to use the exact same inputs, otherwise your comparisons aren't going to be meaningful.

Pavel

On 11/17/19 04:48, Chancievan Thangaratnarajah wrote:

...

Hello,

I have noticed differences in the refinement statistics between two different Phenix versions (see attachments) even though I am using the same cryo-EM map and model as input. The most notable difference are rotamer outliers, which were 0 % in Phenix 1.16-3549 and now 6 % in Phenix 1.17-3660. I was wondering whether anyone has noticed a similar behaviour? Or is there something that I have missed?

Best,

Chancie.