Moorhen's density fit plot (in the Validation plot panel) is giving me some unusual results. Here's an example of a model fitted with Molrep where the last few residues don't fit the map density:
And the same map and model in Coot 0.9:
These residues are basically in the noise so I assume the map density is below zero in places, and Coot 0.9 (correctly, IMO) shows that as a poor density fit while Moorhen shows it with a negative bar instead, which is much harder to interpret.
A few comments and suggestions:
Could Moorhen show poorly-fitting residues with bars that all go in a consistent direction, regardless of whether the map density is above or below zero? (Or use colouring to make it clear that large bars in either direction are both bad!)
It's hard to understand the scales used in the Moorhen validation plot. In Coot 0.9 the red-green colouring makes it very clear that a large bar indicates a poor fit. In Moorhen, I'm not so sure. Do large bars always indicate a problem or are they actually a good thing for some of the metrics? Could you add a guide to interpretation? Either something visual on the plot, or maybe a help panel that the user can open up to explain what the plots mean?
An option to rescale the plots would be helpful, e.g. if there's a large bar that makes all the others so close to zero that they can't be seen
It'd be nice to have a way to view the raw map density values for a particular residue - i.e. take the same map density information that's used internally to calculate the density fit, and make that visible somewhere, maybe as labels on the atoms or entries in a "residue info" table
Moorhen's density fit plot (in the Validation plot panel) is giving me some unusual results. Here's an example of a model fitted with Molrep where the last few residues don't fit the map density:
And the same map and model in Coot 0.9:
These residues are basically in the noise so I assume the map density is below zero in places, and Coot 0.9 (correctly, IMO) shows that as a poor density fit while Moorhen shows it with a negative bar instead, which is much harder to interpret.
A few comments and suggestions: