AlphaFold2 coloring is not according to the database
Hi,
I was preparing a figure and we opened the same AF2 structure both in chimerax and pymol, and colored according to pLDDT. We expected something quite similar, but some of the residues were not colored the same (left chimerax, middle official AF2 coloring, right pymol)
[image: image.png][image: image.png][image: image.png]
These are the corresponding CA atoms from the file, there are several that are above 70. ATOM 2 CA MET B 1 26.541 -48.983 -34.346 1.00 49.18 C ATOM 10 CA ALA B 2 23.562 -47.870 -36.454 1.00 60.35 C ATOM 15 CA ASP B 3 25.250 -49.497 -39.528 1.00 68.46 C ATOM 23 CA ALA B 4 25.752 -52.804 -37.641 1.00 69.86 C ATOM 28 CA TRP B 5 22.024 -53.025 -36.747 1.00 68.46 C ATOM 42 CA GLU B 6 21.036 -52.590 -40.351 1.00 71.02 C ATOM 51 CA GLU B 7 23.249 -55.575 -41.294 1.00 76.21 C ATOM 60 CA ILE B 8 21.578 -57.739 -38.645 1.00 74.81 C ATOM 68 CA ARG B 9 18.177 -56.653 -39.803 1.00 70.07 C ATOM 79 CA ARG B 10 19.226 -57.507 -43.334 1.00 74.00 C ATOM 90 CA LEU B 11 20.598 -60.932 -42.264 1.00 74.36 C ATOM 98 CA ALA B 12 17.459 -61.553 -40.194 1.00 69.47 C ATOM 103 CA ALA B 13 15.384 -60.625 -43.205 1.00 69.30 C ATOM 108 CA ASP B 14 17.563 -62.981 -45.430 1.00 70.53 C ATOM 116 CA PHE B 15 17.273 -65.769 -42.888 1.00 71.78 C ATOM 127 CA GLN B 16 13.568 -65.190 -42.812 1.00 64.74 C ATOM 136 CA ARG B 17 13.352 -65.300 -46.586 1.00 65.36 C ATOM 147 CA ALA B 18 15.491 -68.455 -46.645 1.00 64.05 C ATOM 152 CA GLN B 19 13.261 -70.000 -44.028 1.00 60.46 C ATOM 161 CA PHE B 20 10.117 -69.389 -46.168 1.00 62.26 C ATOM 172 CA ALA B 21 11.642 -70.083 -49.654 1.00 56.65 C ATOM 177 CA GLU B 22 11.244 -73.693 -49.767 1.00 53.68 C ATOM 186 CA ALA B 23 7.622 -74.493 -50.250 1.00 56.47 C
After looking at again the docs, I understood that chimerax has a different coloring [image: image.png] AFDB [image: image.png]
Can you maybe change this? Most people are probably used to the database's coloring, and could interpret it wrong (like me... It drove me crazy when showing somebody else, why the plddt-s do not match to what I have shown before, just in a figure by chimerax). I understand that in the documentation the colorbar is according to the coloring that chimerax does, but it would be better if it was according to the standard.
Thank you so much!
Kind regards, Julia Varga
Hi Julia,
It looks like the difference here is that PyMol is simply colouring the residues in bins (i.e. everything in the same band has exactly the same colour) whereas ChimeraX is using a continuous colour map (i.e. residues in the 50-70 range shade from orange to yellow; 70-90 from yellow to cyan; 90-100 from cyan to deep blue). I'd argue that the latter approach is better since it gives you a much more fine-grained idea of confidence - in the PyMol scheme, for example, a residue with a pLDDT of 70.1 will look much more confident than one with a pLDDT of 69.9, which isn't really ideal.
If you're absolutely sure that's what you want, you can get the "binned" effect with something like:
alias afcolor color bfactor $1 palette 0,orange:49.9,orange:50,yellow:69.9,yellow:70,cyan:89.9,cyan:90,blue afcolor #1
Best, Tristan
On Mon, Jun 5, 2023 at 12:09 PM Julia Varga via ChimeraX-users < chimerax-users@cgl.ucsf.edu> wrote:
Hi,
I was preparing a figure and we opened the same AF2 structure both in chimerax and pymol, and colored according to pLDDT. We expected something quite similar, but some of the residues were not colored the same (left chimerax, middle official AF2 coloring, right pymol)
[image: image.png][image: image.png][image: image.png]
These are the corresponding CA atoms from the file, there are several that are above 70. ATOM 2 CA MET B 1 26.541 -48.983 -34.346 1.00 49.18 C ATOM 10 CA ALA B 2 23.562 -47.870 -36.454 1.00 60.35 C ATOM 15 CA ASP B 3 25.250 -49.497 -39.528 1.00 68.46 C ATOM 23 CA ALA B 4 25.752 -52.804 -37.641 1.00 69.86 C ATOM 28 CA TRP B 5 22.024 -53.025 -36.747 1.00 68.46 C ATOM 42 CA GLU B 6 21.036 -52.590 -40.351 1.00 71.02 C ATOM 51 CA GLU B 7 23.249 -55.575 -41.294 1.00 76.21 C ATOM 60 CA ILE B 8 21.578 -57.739 -38.645 1.00 74.81 C ATOM 68 CA ARG B 9 18.177 -56.653 -39.803 1.00 70.07 C ATOM 79 CA ARG B 10 19.226 -57.507 -43.334 1.00 74.00 C ATOM 90 CA LEU B 11 20.598 -60.932 -42.264 1.00 74.36 C ATOM 98 CA ALA B 12 17.459 -61.553 -40.194 1.00 69.47 C ATOM 103 CA ALA B 13 15.384 -60.625 -43.205 1.00 69.30 C ATOM 108 CA ASP B 14 17.563 -62.981 -45.430 1.00 70.53 C ATOM 116 CA PHE B 15 17.273 -65.769 -42.888 1.00 71.78 C ATOM 127 CA GLN B 16 13.568 -65.190 -42.812 1.00 64.74 C ATOM 136 CA ARG B 17 13.352 -65.300 -46.586 1.00 65.36 C ATOM 147 CA ALA B 18 15.491 -68.455 -46.645 1.00 64.05 C ATOM 152 CA GLN B 19 13.261 -70.000 -44.028 1.00 60.46 C ATOM 161 CA PHE B 20 10.117 -69.389 -46.168 1.00 62.26 C ATOM 172 CA ALA B 21 11.642 -70.083 -49.654 1.00 56.65 C ATOM 177 CA GLU B 22 11.244 -73.693 -49.767 1.00 53.68 C ATOM 186 CA ALA B 23 7.622 -74.493 -50.250 1.00 56.47 C
After looking at again the docs, I understood that chimerax has a different coloring [image: image.png] AFDB [image: image.png]
Can you maybe change this? Most people are probably used to the database's coloring, and could interpret it wrong (like me... It drove me crazy when showing somebody else, why the plddt-s do not match to what I have shown before, just in a figure by chimerax). I understand that in the documentation the colorbar is according to the coloring that chimerax does, but it would be better if it was according to the standard.
Thank you so much!
Kind regards, Julia Varga
ChimeraX-users mailing list -- chimerax-users@cgl.ucsf.edu To unsubscribe send an email to chimerax-users-leave@cgl.ucsf.edu Manage subscription:
participants (2)
-
Julia Varga -
Tristan Croll