How to make the 2D result from crySPARC a template for autopicking in Relion?

Hi all,

I have some trouble in analyzing negative stain micrographs.

  1. I manually picked ~1000 particles for 2D classification. But when I tried to use the selected 2D classes for template picking, it seems like Cryosparc picked the dark background instead of the white particles. I know this is how it works for autopicking particles from Cryo-images in which the particle is black. But is there any solution for the negative stain?

  2. When I tried to use the generated .mrc file to serve as a template for autopicking in relion, it seems like I need to convert the .mrc to .mrcs. How could I achieve that?

  3. Where can I find the information about the pixel size of the generated .mrc file by cryoSPARC?


  1. Not sure about this one, but when you import your micrograph to cryoSPARC, you should tell what type of images you are importing (negative staining, black-on-white).

  2. Just rename it to .mrcs

  3. You can use “header” command from IMOD to display the pixel space information of the mrc file.

1 Like

Hi Bing,

Thanks for the answer! But could you show me how to tell cryoSPARC that the image is negative staining?

Probably you should tell cryoSPARC when you extract your particles. see the last option:

I opt to simply invert my negative stain micrographs using (EMAN2): *.DM4 @.mrc --mult=-1

Disclaimer: this is for quickly checking if the protein is good or not, not for publication.

1 Like

Hi @ruiruigo, @bing, @ZhijieLi, @DanielAsarnow,

For now, we recommend the solution posted above and here: Easy template picking

We’re aware of the inconsistency with how negative stain data is handled and will work to fix it in an upcoming release.


1 Like