I've been trying to figure this out for like 2 days. As someone who's never used PyMOL since this week, this is incredibly clear. Thank you so much, seriously :')
Thanks, great content! Btw, do you have any idea why is the electron density map called "2 fofc"? I don't understand the origin of the letters "fofc", that is, google doesn't seem to help either.. edit, answer: "It is called 2Fo-Fc because the calculation involves combining the observed diffraction data, Fo, with the diffraction data calculated from the atomic model, Fc, in a way that gives the least-biased result"
Fo refers to the experimentally resolved data and Fc refers to the expected density based on the model you provide. The 2FoFc map is therefore weighted toward the experimental data rather than the model data (proteopedia.org/wiki/index.php/Electron_density_maps). This is important to know when resolving structures by crystallography as the map density is influenced by the presence of model to which it has been refined (the R-value, aka R-work), which means that it is possible to 'overfit' the density to model. This is overcome by taking a portion of the data which remains uninfluenced by the model (R-free) which can then be compared to the model independently to the fitting (pdb101.rcsb.org/learn/guide-to-understanding-pdb-data/r-value-and-r-free). Hope that helps.
@@pymolross Thank you! Now found the abbreviation origin also - I found the formula from a pdf called "Interpretation of electron density" from google: Fo = structure Factor (amplitudes) Observed, Fc = structure Factor (amplitudes) Calculated; in the formula, 2 FoFc has a factor of 2 for the Fo (=Fobs), whereas in FoFc there is a factor of 1. I wonder if this just means that Fo gets more emphasis in 2FoFc. I had missed the Fo and Fc during the course I'm taking, I just remembered R-values roughly :D There's so many methods (X-ray, cryo-em, NMR, and MD we went through during the course) and not being a mathematician these are often over-the-head level for me. Thank you, I already had forgotten about the model fitting altogether.
Yes, so the 2FoFc map is weighted towards the Fo, however the FoFc map is equally weighted, which is why we call the FoFc map the 'difference' map, as when contoured to a sigma value of +3 or -3 it highlights discontinuities between the observed density and the calculated density.
Hi Anytimeisok, the command you want is 'isomesh' (pymolwiki.org/index.php/Isomesh) First load in your map, then follow the examples on the web page above, let me know how you get on.
@@pymolross Hi Thanks for the tip and I have tried with a PDB from PDB database and it works. However I failed when I tried with my map and model (the density does not fit with the ligand position eventhough I have fit the model (.pdb) into the map (.mrc)). I am wondering the map is .mrc and it is not the same as 2fofc?
2FoFc refers to the weighting of the map between the experimental data (Fo) and input model (Fc), the map you have in .mrc is a format i am unfamiliar with i'm afraid. So to confirm, the protein does fit the density, but the ligand density is not present? If this is the case then I'm afraid there is not experimental evidence of your ligand interacting with your protein in this experiment. In case i misinterpreted your question, here are some other ideas . . . When you load the map and generate an isomesh, you do get the map? If you do not get a map here either 1. The map file type is not supported by pymol (i would presume this is not the case) 2. You may need to change the level of the map, try 0.5,1, 2, 3, (I have not fit EM maps, i presume they are scaled to sigma (stdev) as in x-ray diffraciton) Does the map fit with to the protein, but not the ligand? If the protein does not fit to the map, Then make sure you are using the output coordinates (pdb) generated from the fitting program. If the protein fits, but the ligand is not present, then there is not sufficient occupancy of the ligand in the binding site. This is equivalent to an unbound structure, It is possible that you can reduce sigma to 1.0 to see a low occupancy ligand, you can even check in 0.5 sigma, but note, that there will be significant noise here. Best to work at sigma 2.0
@@pymolross Thanks for the information. I also checked the map and the model in Coot and I do see the density for the ligand. I will play around with sigma and see how it goes. Thanks again!
Hi @rafffa26, I'm pretty sure that you can just open ccp4 maps (for example) from the standard 'open' menu, you have to export the map from ccp4 in the correct format though. this may help pymolwiki.org/index.php/Display_CCP4_Maps
@@pymolross I used this link, I can open the ccp4 maps but I cannot superimpose it/align the electron density to the PDB. I already tried doing the super, and align commands and I get errors.
@@nicolefanabrito7028 So, to confirm, you are trying to align a pdb and 'homemade map' ie one not from the pdb? Any maps from the pdb should be pre-aligned Any homemade maps should be aligned to the molecules it was used to generate, so long as you have saved the coordinates. You cannot align molecules and maps as the alignment is based on an RMSD between the atoms in both files (the map of course has no atoms). But what you can do is transform the object (pymolwiki.org/index.php/Transform_selection) or instead you can perform separate rotations and translations (pymolwiki.org/index.php/Rotate, pymolwiki.org/index.php/Translate), provided you know the transformation matrix. If none of that is helpful, can you provide more details of what you are starting with. Thanks
I've been trying to figure this out for like 2 days. As someone who's never used PyMOL since this week, this is incredibly clear. Thank you so much, seriously :')
Glad it helped!
Thanks, great content! Btw, do you have any idea why is the electron density map called "2 fofc"? I don't understand the origin of the letters "fofc", that is, google doesn't seem to help either..
edit, answer: "It is called 2Fo-Fc because the calculation involves combining the observed diffraction data, Fo, with the diffraction data calculated from the atomic model, Fc, in a way that gives the least-biased result"
Fo refers to the experimentally resolved data and Fc refers to the expected density based on the model you provide. The 2FoFc map is therefore weighted toward the experimental data rather than the model data (proteopedia.org/wiki/index.php/Electron_density_maps). This is important to know when resolving structures by crystallography as the map density is influenced by the presence of model to which it has been refined (the R-value, aka R-work), which means that it is possible to 'overfit' the density to model. This is overcome by taking a portion of the data which remains uninfluenced by the model (R-free) which can then be compared to the model independently to the fitting (pdb101.rcsb.org/learn/guide-to-understanding-pdb-data/r-value-and-r-free).
Hope that helps.
@@pymolross Thank you! Now found the abbreviation origin also - I found the formula from a pdf called "Interpretation of electron density" from google: Fo = structure Factor (amplitudes) Observed, Fc = structure Factor (amplitudes) Calculated; in the formula, 2 FoFc has a factor of 2 for the Fo (=Fobs), whereas in FoFc there is a factor of 1. I wonder if this just means that Fo gets more emphasis in 2FoFc.
I had missed the Fo and Fc during the course I'm taking, I just remembered R-values roughly :D There's so many methods (X-ray, cryo-em, NMR, and MD we went through during the course) and not being a mathematician these are often over-the-head level for me.
Thank you, I already had forgotten about the model fitting altogether.
Yes, so the 2FoFc map is weighted towards the Fo, however the FoFc map is equally weighted, which is why we call the FoFc map the 'difference' map, as when contoured to a sigma value of +3 or -3 it highlights discontinuities between the observed density and the calculated density.
I have been trying to find how to present the mesh for cryoEM dataset. Could you please give me some tips? Thanks!
Hi Anytimeisok,
the command you want is 'isomesh' (pymolwiki.org/index.php/Isomesh)
First load in your map, then follow the examples on the web page above, let me know how you get on.
@@pymolross Hi Thanks for the tip and I have tried with a PDB from PDB database and it works. However I failed when I tried with my map and model (the density does not fit with the ligand position eventhough I have fit the model (.pdb) into the map (.mrc)). I am wondering the map is .mrc and it is not the same as 2fofc?
2FoFc refers to the weighting of the map between the experimental data (Fo) and input model (Fc), the map you have in .mrc is a format i am unfamiliar with i'm afraid.
So to confirm, the protein does fit the density, but the ligand density is not present?
If this is the case then I'm afraid there is not experimental evidence of your ligand interacting with your protein in this experiment.
In case i misinterpreted your question, here are some other ideas . . .
When you load the map and generate an isomesh, you do get the map?
If you do not get a map here either
1. The map file type is not supported by pymol (i would presume this is not the case)
2. You may need to change the level of the map, try 0.5,1, 2, 3, (I have not fit EM maps, i presume they are scaled to sigma (stdev) as in x-ray diffraciton)
Does the map fit with to the protein, but not the ligand?
If the protein does not fit to the map,
Then make sure you are using the output coordinates (pdb) generated from the fitting program.
If the protein fits, but the ligand is not present, then there is not sufficient occupancy of the ligand in the binding site.
This is equivalent to an unbound structure, It is possible that you can reduce sigma to 1.0 to see a low occupancy ligand, you can even check in 0.5 sigma, but note, that there will be significant noise here. Best to work at sigma 2.0
@@pymolross Thanks for the information. I also checked the map and the model in Coot and I do see the density for the ligand. I will play around with sigma and see how it goes. Thanks again!
How do you open an electron density map from your computer, not deposited yet? Many thanks.
Hi @rafffa26, I'm pretty sure that you can just open ccp4 maps (for example) from the standard 'open' menu, you have to export the map from ccp4 in the correct format though. this may help pymolwiki.org/index.php/Display_CCP4_Maps
@@pymolross I used this link, I can open the ccp4 maps but I cannot superimpose it/align the electron density to the PDB. I already tried doing the super, and align commands and I get errors.
@@nicolefanabrito7028
So, to confirm, you are trying to align a pdb and 'homemade map' ie one not from the pdb?
Any maps from the pdb should be pre-aligned
Any homemade maps should be aligned to the molecules it was used to generate, so long as you have saved the coordinates.
You cannot align molecules and maps as the alignment is based on an RMSD between the atoms in both files (the map of course has no atoms). But what you can do is transform the object (pymolwiki.org/index.php/Transform_selection) or instead you can perform separate rotations and translations (pymolwiki.org/index.php/Rotate, pymolwiki.org/index.php/Translate), provided you know the transformation matrix.
If none of that is helpful, can you provide more details of what you are starting with.
Thanks