[phenixbb] phenix.model_vs_data

Nathaniel Echols nechols at lbl.gov
Thu May 13 10:50:35 PDT 2010


On Thu, May 13, 2010 at 10:37 AM, Young-Jin Cho <yjcho at brandeis.edu> wrote:

> Dear Phenix users,
> This might be trivial questions though...
> 1. When you did phenix.model_vs_data, I see the discrepancy between
> R/Rfree.
>  'After applying resolution and sigma cutoffs' seems more accurate one
> though, could someone explain why it has such-even small though-
> discrepancy?
>

Not sure I understand the question - what kind of discrepancy?  R-free
should always be a little higher than R-work...

2. Also, when you do phenix.polygon, I used to seeing many branches from old
> version. With my current version(phenix-1.6.1-343), I normally see only 5
> branches. Is there any reasons for that?
>

Yes, we decided that the additional statistics were potentially too
confusing and less appropriate as global measures of structure quality (in
particular, k_sol should never have been included).  You can still add more
if you want by clicking the button labeled "Select statistics for display".

PS. You should update to the latest official release - 1.6.1-343 was an
unofficial nightly build.  (Or better yet, just use the latest nightly
build, which has many improvements over 1.6.1.)

-Nat
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phenix-online.org/pipermail/phenixbb/attachments/20100513/53edc2bc/attachment-0003.htm>


More information about the phenixbb mailing list