From: Xiaopeng Hu
Date: 2011/9/30
Dear all,
I just notified that there is a big difference between the Ramachandran plot analysis results produced by Coot and Morprobity (or Phenix). For the structure I am working now, Phenix(Morprobity) gives out Ramachandran outliers 0.2%, favored 95.2%, whileas Coot gives out Outliers 1.24%, Allowed 5,14% and Prefered 93.62%.I am wondering if there is a simply explain for the difference which I don't know? Or I just made some silly mistakes?
Best wishes,
xiaopeng
----------
From: Edward A. Berry
What does ProCheck say?
----------
From: Frank von Delft
Oh god no don't ask Procheck , its Rama plot is a complete disaster zone - for one thing, it's * ancient*. Doesn't discriminate between amino avoid types. Grrr. To be avoided at all costs.
Sent from tiny silly touch screen
----------
From: Mario Sanches
Hi Xiaopeng,
----------
From: Nat Echols
Date: 2011/9/30
Dear all,
I just notified that there is a big difference between the Ramachandran plot analysis results produced by Coot and Morprobity (or Phenix). For the structure I am working now, Phenix(Morprobity) gives out Ramachandran outliers 0.2%, favored 95.2%, whileas Coot gives out Outliers 1.24%, Allowed 5,14% and Prefered 93.62%.I am wondering if there is a simply explain for the difference which I don't know? Or I just made some silly mistakes?
Best wishes,
xiaopeng
----------
From: Edward A. Berry
What does ProCheck say?
----------
From: Frank von Delft
Oh god no don't ask Procheck , its Rama plot is a complete disaster zone - for one thing, it's * ancient*. Doesn't discriminate between amino avoid types. Grrr. To be avoided at all costs.
Sent from tiny silly touch screen
----------
From: Mario Sanches
Hi Xiaopeng,
If you are feeding both programs exactly the same file then you are not doing anything wrong. Notice that, if you calculate the Ramachandran plot on Molprobity and then you go on coot, do a round of manual refinement, and only then calculate the Ramachandran, then that can be the source of the difference. But I would bet that it is because different programs have their own definition of what is allowed and what is an outlier.
I personally use the Ramachandran plot on coot just to guide my refinement, but use another program (Molprobity is pretty good) to do a thorough validation.
Good luck,
Mario Sanches
--
Mario Sanches
Postdoctoral Researcher
Samuel Lunenfeld Research Institute
Mount Sinai Hospital
600 University Ave
Toronto - Ontario
Canada
M5G 1X5
http://ca.linkedin.com/in/mariosanches
Mario Sanches
Postdoctoral Researcher
Samuel Lunenfeld Research Institute
Mount Sinai Hospital
600 University Ave
Toronto - Ontario
Canada
M5G 1X5
http://ca.linkedin.com/in/mariosanches
----------
From: Nat Echols
It's not you - I believe that Coot and Molprobity/Phenix both use the Richardson lab's data, but this isn't the first complaint I've heard about discrepancies in the statistics, so I suspect that the cutoffs are slightly stricter in Coot. Will check this later today.
-Nat
No comments:
Post a Comment