AlignProjections: Align projections of PDB/MRC volumes with class averages

This program will align a PDB file or a 3D MRC with a 2D class average stack (.mrcs). Aligning projections with class averages serve as a validation of your 3D reconstruction. OR, this will tell you if your PDB model agrees with your class averages.

Why should you use this?

When you want to align a PDB or MRC volume with a stack of class averages (or projections in a stack).

E.g., let’s say you have these class averages (.mrcs format) and you’d like to know if it matches a 3D reconstruction:

(These are beta-galactosidase class averages from RELION tutorial)

How does this program run?

This program will use SPIDER for the following steps:

  1. Forward project input 3D map
  2. Align each class average with the best matching projection
  3. Display class averages alongside best matching projection (rotated in the registry)

Aligning PDB file:

If you’re aligning a PDB file, the program will convert the atomic model into density using (from EMAN2) first, then align projections with class averages.

How to run this program on COSMIC²

Input files

  • Input for job > Upload a PDB file OR a 3D volume using web browser upload
    • Important: If you upload a 3D volume, the volume must have the same pixel size and dimensions as the class averages.
  • Selected when creating a task > Upload a stack of class averages from any software (RELION, cryoSPARC, etc) that are in .mrcs format using web browser upload
    • You can download the class averages from cryoSPARC by clicking the download button for class averages > Download outputs > ‘Blob’


  • Resolution limit for PDB upload – this resolution is used to filter the converted PDB to MRC volume
  • (Advanced) Angular sampling – this specifies how finely the projections will be sampled from the 3D volume
  • (Advanced) Pixel size of class averages – input pixel size here, otherwise the program reads the pixel size from the header
  • (Advanced) B-factor to apply to converted PDB file – a positive value here will also act like a low pass filter to help the atomic model agree with your averages in appearance

Example output

The program will output the class averages shown alongside projections as .mrcs or .png file:


The output stack will output images in pairs:

class average #1 + matching projection #1 ;

class average #2 + matching projection #2 ;