Dear Experts, I have a map generated through cryosparc (ab initio-reconstr and homo-refine at 3.7A* FSC and sharpened in Scipion and separately in EMAN2). I have build the initial pdb through BALBES which has perfect validation parameters and nice fitting in the map (with the mtz output) but the co-ordinates are shifted from the true origin of CS-map. I used phenix-RSR to fit and refine, but validation results remained too far to accept the model as good solution and also a small region entered in wrong nearyby densities. So far following independent and downstream approaches I have taken:- In phenix; dock in map, map to model, rigid body refine followed by real space refine (with both .map and converted .mtz), using only chain A-map boxing and model generation through NCS. I have tried chimera fit for entire assembly and also only one chain in boxed map and full map but still the validation remained far from acceptable values and a small region in wrong neighboring density. Please help. -- Best regards, Promod
Hi Pramod, I need to see the map and models before and after refinement to diagnose the issue. Can you share the files? I will have a look once I have the files. Pavel On 10/22/19 10:34, pramod kumar wrote:
Dear Experts,
I have a map generated through cryosparc (ab initio-reconstr and homo-refine at 3.7A* FSC and sharpened in Scipion and separately in EMAN2). I have build the initial pdb through BALBES which has perfect validation parameters and nice fitting in the map (with the mtz output) but the co-ordinates are shifted from the true origin of CS-map. I used phenix-RSR to fit and refine, but validation results remained too far to accept the model as good solution and also a small region entered in wrong nearyby densities.
So far following independent and downstream approaches I have taken:-
In phenix; dock in map, map to model, rigid body refine followed by real space refine (with both .map and converted .mtz), using only chain A-map boxing and model generation through NCS.
I have tried chimera fit for entire assembly and also only one chain in boxed map and full map but still the validation remained far from acceptable values and a small region in wrong neighboring density.
Please help.
-- Best regards, Promod
Dear Pavel,
Thank you so much for the consideration, I have shared the file with you.
Best regards,
Pramod
On Tue, Oct 22, 2019 at 7:26 PM Pavel Afonine
Hi Pramod,
I need to see the map and models before and after refinement to diagnose the issue. Can you share the files? I will have a look once I have the files.
Pavel
On 10/22/19 10:34, pramod kumar wrote:
Dear Experts,
I have a map generated through cryosparc (ab initio-reconstr and homo-refine at 3.7A* FSC and sharpened in Scipion and separately in EMAN2). I have build the initial pdb through BALBES which has perfect validation parameters and nice fitting in the map (with the mtz output) but the co-ordinates are shifted from the true origin of CS-map. I used phenix-RSR to fit and refine, but validation results remained too far to accept the model as good solution and also a small region entered in wrong nearyby densities.
So far following independent and downstream approaches I have taken:-
In phenix; dock in map, map to model, rigid body refine followed by real space refine (with both .map and converted .mtz), using only chain A-map boxing and model generation through NCS.
I have tried chimera fit for entire assembly and also only one chain in boxed map and full map but still the validation remained far from acceptable values and a small region in wrong neighboring density.
Please help.
-- Best regards, Promod
participants (2)
-
Pavel Afonine -
pramod kumar