What you want to do is to find the principal axes of your object. you can use this: https://github.com/pierrepo/principal_axes

In PyMol, the command "orient" https://pymolwiki.org/index.php/Orient aligns the longest axis to the x-axis of the view, the second-longest to the y-axes and the shortest to the z-axis (perpendicular to the plane of the image). You can then use the measurement wizard to determine the distance between the atoms at the extreme ends. I personally like to using the builder to generate a (G4AG4K)n beta strand as an internal ruler.

You can also use the "bounding box" https://pymolwiki.org/index.php/Bounding_Box or "draw bounding box" https://pymolwiki.org/index.php/DrawBoundingBox scripts to determine the principal axes and draw a bounding box around your molecule. DrawBoundingBox will also print the box dimensions

In VMD: https://www.ks.uiuc.edu/Research/vmd/script_library/scripts/orient/

In Modeller: The MODELLER software package has a built-in orient command that is very fast: http://salilab.org/modeller/9v4/manual/node180.html For you, if you don't want to actually orient it, just skip the mdl.write() line and use the r.translation() data. Note that since MODELLER is written for homology stuff this will ignore hydrogens by default. To keep your hydrogens you'll want to add this line right after the env = environ() line:env.io.hydrogen = True

In ChimeraX: https://www.cgl.ucsf.edu/chimerax/docs/user/tools/axesplanes.html

As to the "correctness" of the results: For a single, well folded domain, the results for different structures of the same protein will agree very well. For multi-domain proteins in which the different domains are connected by flexible hinges or linkers, different conformations will of course result in different dimensions (e.g. the different full-length IgG structures in the PDB)

Annemarie Honegger Thank you so much!