Hi Nathan, The error message "No alphafold specified for alphafold contacts" is goofed up -- it should say "No atoms specified for alphafold contacts". This means your structure has no chain A, so your command "alphafold contacts /A to /B distance 5 maxpae 6" needs to be modified to use the actual chain identifiers in your structure. You can hover the mouse over the chains in your structure and it will popup some text indicating the chain id under the mouse. Yes the "Alphafold Error Plot" in ChimeraX works for Boltz 1 and 2, AlphaFold 2 and 3, OpenFold 3, and ESMFold predictions. I plan to rename it Predicted Aligned Error Plot so it will be clearer that it is not specific to AlphaFold. It currently has the AlphaFold name because when I added AlphaFold 2 support about 4 years ago there were no other prediction programs writing PAE. I have fixed the text of the error message in the daily build. Thanks for reporting the problem. Tom
On Apr 21, 2026, at 4:41 PM, Nathan Davis <ndavis18@uwyo.edu> wrote:
I get the error: No alphafold specified for alphafold contacts. I attached my chimerax session. Should I be using alphafold error plot for the pae?
On Apr 21, 2026, at 11:06 AM, Nathan Davis via ChimeraX-users <chimerax-users@cgl.ucsf.edu <mailto:chimerax-users@cgl.ucsf.edu>> wrote:
Hi, I'm Nathan. I have been using ChimeraX to assess protein-protein interfaces. One of the commands I use is alphafold contacts, usually with something like this: alphafold contacts /A to /B distance 5 maxpae 6. I started looking at boltz runs as well and realized that this command does not work for boltz models. Is there a similar command that would work for boltz? Also, I'm importing boltz models that I ran on my local cluster.
Thank you! Nathan