Alphafold 3 introduced PAE values for individual ligand atoms. Alphafold 2 only predicted proteins, not ligands, and its PAE values were all per-residue. That is why ChimeraX now handles PAE values either per-residue or per-atom so it can handle both AF3 and AF2 PAE output.

The PAE heatmap display in ChimeraX just displays the values in the AF3 or AF2 output. It hasn't changed in any significant way. So the difference you are seeing is probably because your AF3 predictions have very different PAE values thatn the AF2 predictions you did before. You could test this by looking at your older AF2 predictions in the current ChimeraX and I think you will see it gives the results you saw a few years ago with AF2.

AF3 is not just a refined version of AF2, it is is a totally new program that uses new methods from predicting structures. So you should not expect the results to be the same for the two programs, especially if the predictions have borderline confidence values.

On Mar 8, 2026, at 8:28 AM, Rowan Cain via ChimeraX-users <chimerax-users@cgl.ucsf.edu> wrote:

Hello,

I am an undergraduate doing research involving protein protein interactions in C. elegans, and I have recently been having some issues with the alphafold contacts command when measuring buried area in protein-protein interfaces. I am running the same batch of dimers that was measured a few years ago with AF2 data, and now with current AF3 data I am seeing vastly different results with the same maxPae, distance values. The main change I noticed with ChimeraX is that the argument in the alphafold contacts command has been changed from “toResidues” to “toAtoms” - is there any reason why this change took place, and would it cause the amount of interacting residues to be measured differently?
Thank you,
Rowan Cain

_______________________________________________
ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu
To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu
Archives: https://mail.cgl.ucsf.edu/mailman/archives/list/chimerax-users@cgl.ucsf.edu/