Programmatic access to cryosparc

Features and Functionality
1

Hi, if i already have aligned movies (micrographs) and ctf estimations done externally from cryosparc, is there a programmatic (ie API) way of registering the micrographs and ctf results into a new workflow? thanks,

2 Likes
2

Agreed, this would be very desirable. Some sort of import motioncorr2/GCTF job types?

3

I agree - this should already be in the Import Micrographs job - I should be able to provide a star file with micrographs and ctf parameters, and start from picking. It is puzzling that it is possible to import a particle star file with CTF information, but not a micrograph star file.

This would make cryosparc 2 much more useful. E.g. I like doing motion correction in motioncor2 and ctf correction using gctf, but I can’t do that if I want to pick in cryosparc (and I really like the picking in cryosparc). So I don’t pick in cryosparc, I pick in relion and import after that, but if I could just import a micrograph star file it would make everything easier.

Cheers
Oli

4

Agree +1. Because of some reason, motion correction in cryosparc on our cluster is pretty slow. If I could import information from MotionCor2, it would save me a lot of time. I really like to do per particle motion correction in cryosparc!

-Xing

5

there are two things that i would like:

  1. the ability to import pre-ctf’d micrographs (my preference is ctffind, but gctf would be good too) - for example, importing in pre-motion corrected images is quick, but having to run the ctfs again can take 15 mins+.

  2. the ability to use an API to automatically create a new workspace with the above ctf’s already imported.

i agree with @olibclarke and starting form a star file would be probably the most flexible way.

6

Hi,

I found that one can make the CTF job run much faster by turning “Slower, more exhaustive search? [yes]” to [no].

The help says “Answer yes if you expect very high astigmatism (say, greater than 1000A) or in tricky cases. In that case, a slower exhaustive search against 2D spectra (rather than 1D radial averages) will be used for the initial search.”

Zhijie

7

Exhaustive search is a lot slower, but it also works better and more micrographs will have higher CTF resolutions and thus you end up with more data if you filter by CTF fit quality at some point. Give it a try - “no” is the default in cisTEM, by the authors of CTFFIND - but do compare and/or re-run with “yes” once you have a good structure and enter the resolution pushing phase.