|
Size: 1341
Comment:
|
← Revision 3 as of 2021-07-21 20:19:41 ⇥
Size: 572
Comment:
|
| Deletions are marked like this. | Additions are marked like this. |
| Line 3: | Line 3: |
| A: If you are using a recent version of EMAN2, it should be possible to go back and change the box size, but doing so may be somewhat messy, since as you work back forward, you will be displacing some, but not all information from the earlier box size, and invalidating many of your earlier results. For example, the structure factor computed at one box size will have a different sampling than at another box size. Most parameters can easily be changed during a project, but box size becomes quite messy, so, while it may not be required, I generally suggest to people that if they want to change the box size they start over with a fresh project, by copying over the particles from the first project with the new box size. If you wish to copy particles from one project to another with a different box size, I would suggest exporting the particles from the first project as hdf files while shrinking, then importing these into the new empty project. For example, in BASH: for i in `e2bdb.py particles --filtexclude=ctf` do e2proc2d.py bdb:particles#$i <path to new proj>/$i.hdf --clip=<newsize>,<newsize> done then run e2workflow from <path to new proj> and import the hdf particle files (after which you can delete the hdf's) |
A: You don't need to redo particle picking, and you can use 3D maps generated at the original size as initial models for the new size. You need to: * return to the "generate output step" in particle picking and rerun this with the new box size * rerun automatic CTF correction * you should not need to rerun 2-D class averaging or initial model generation as long as you got a reasonable initial model at the old size * run new refinements with the newly reboxed particles |
Q: How can I reduce the box size in a project without starting completely from scratch
A: You don't need to redo particle picking, and you can use 3D maps generated at the original size as initial models for the new size. You need to:
- return to the "generate output step" in particle picking and rerun this with the new box size
- rerun automatic CTF correction
- you should not need to rerun 2-D class averaging or initial model generation as long as you got a reasonable initial model at the old size
- run new refinements with the newly reboxed particles