Re: [scipion-users] Per-particle defocus estimation on tilted micrographs

[Pablo C. <pc...@cn...>]

Hi Teige, I'm very happy to see your enthusiasm, thanks.

As far as I know, unless there are some details I might be missing, you 
currently have (scipion3):

gctf, relion, and cryosparc (in beta)


Regarding Scipion not using gctf-1.06, default values point to 1.18.

You may need:

1.- Install 1.0 binaries: *scipion3 installb gctf-1.06*

2.- Edit/create scipion-conf file:


     If you want to use a different binary then you need to have a 
scipion.conf file (optional and missing by default).

     Something like this placed on <SCIPION_HOME>/config/scipion.conf 
should be the minimal you need:

-------------------scipion.conf file-----------------

[PLUGINS]

GCTF_HOME = /home/pablo/software/scipion/software/em/gctf-1.06
GCTF = Gctf-v1.06_sm_30_cu8.0_x86_64
GCTF_CUDA_LIB = /usr/local/cuda-8.0/lib64

---------------------------------------------------------

NOTE: Adapt the paths to your SCIPION_HOME and cuda version you want to use.


To know which variables can be set for a specific plugin try:

 >> *scipion3 config -p gctf*

Scipion v3.0.5 - Eugenius
Variables defined by plugin 'gctf':

GCTF_HOME = /home/pablo/software/scipion/software/em/gctf-1.18
GCTF = Gctf_v1.18_sm30-75_cu10.1
GCTF_CUDA_LIB = /usr/local/cuda/lib64

These variables can be added/edited in 
'/home/pablo/software/scipion/config/scipion.conf'


Each plugin has it's small documentation page where these details are 
usually commented:
See https://github.com/scipion-em/scipion-em-gctf


Cheers, Pablo.

On 6/11/20 7:54, Matthews-Palmer, Teige wrote:
> Dear Scipion users & devs,
>
> I’m very excited to be installing Scipion v3.0.5 at a new job - it has been a while since I used Scipion and very many programs and features have been added, it looks wonderful.
>
> One feature I would really like to use in Scipion is accurate CTF estimation for particles on tilted micrographs.
> As far as I know the options are patch-based CTF in gCTF (initially, before particle picking), per-particle in gCTF, geometry aware CTF in goCTF, and something similar in Warp. Maybe there are other options that I’m not aware of - I’d be happy to hear about them!
>
> It looks like local or per-particle CTF are not available with gCTF in Scipion3 - is that right?
>
> Using the plugin manager I installed gctf-1.18
> Then, since the github page mentions 1.18 is specifically for phase plate, I installed gctf-1.06 with the plugin manager - but the ProtGctf protocol in Scipion is not changed and still has no options relating to patch or per-particle CTF.
>
> So, can I get accurate particle defocus values from tilted data in a Scipion workflow?
>
> Many thanks,
> Teige
> _______________________________________________
> scipion-users mailing list
> sci...@li...
> https://lists.sourceforge.net/lists/listinfo/scipion-users
-- 
Pablo Conesa - *Madrid Scipion <http://scipion.i2pc.es> team*