FreAlign in EMAN2

Forums

Hi Niko,

I have a medium resolution structure of a virus particle which I got using EMAN2 (e2refine_easy). I tried to do a refinement using FreAlign in EMAN2. Unfortunately, the "refined" structure looks less detailed than the input (snapshot attached - input on the left, OutputMap on the right). I am most likely using it wrong/inputting some wrong parameters - where could I be going wrong? Can you please give me some suggestions on how to improve the FreAlign refinement results?

- Convert to FreAlign
e2refinetofrealign.py refine_16 4 --mode=1 --fbeaut --reslow=100.0 --reshigh=10.0 --rrec=13.6 --rclas=8.0 --mass=9000.0 --interp=0 --imem=1 --verbose=1

- Run FreAlign 3D refinement
e2runfrealign.py frealign_06

- Covert from FreAlign to EMAN2
e2refinefromfrealign.py frealign_06 --icosahedral_symmetry

- EMAN version
EMAN 2.1 (CVS 2014/11/11 17:17:45)
Your EMAN2 is running on: Linux-2.6.32-504.el6.x86_64-x86_64-with-redhat-6.6-Santiago 2.6.32-504.el6.x86_64 x86_64
Your Python version is: 2.7.3

- FreAlign version
frealign_v9.07_140328

(On a side note, I also tried the same project in parallel in Scipion and used FreAlign for refinement of a structure that I had obtained using Relion 3D refinement. In that project too, it seemed the refined output was less detailed - almost like a featureless ball. But I read here that someone had a similar problem - I will have to try many more parameter changes to try and troubleshoot it myself. But in case there's something fundamental to keep in mind, please let me know)

Thank you very much.
Saumya

PhD Student
Indian Institute of Technology Delhi
India

It looks like you are not using the latest version of Frealign. Also, I am not familiar with the EMAN2 scripts you are using. Please download the latest version from our web page and run your refinement with this version. As a general note, it often happens that the alignment parameters imported from another software are not correctly translated into Frealign parameters. Provided magnification and defocus parameters are correctly translated, you can run Frealign's global search (Mode 3) to redetermine the alignment parameters. However, this will take considerably more computational time than running in Mode 1.