P K wrote:
I am new to the filed of crystallography. I am having trouble figuring
out what exactly does sigma level of electron density map mean.
When sigma level of a map is increased (say from 1.5 sigma to 2 sigma)
why the map covering individual residues becomes less wide and more
"precise"? Shouldn't it be the other way round if they have anything to
do with the Gaussian districution?
Dear Great Biologist,
Imagine starting on the plains of Tanzania and walking up Mt
Kilimanjaro. Every 100 steps you make an note of your elevation. You
get to the top and make your last note. You come down again. You
decide to do some statistics on your elevation data (relative to you
starting position). You calculate a mean and standard deviation.
Questions the you might like to ask yourself at this point: do the
observations correspond to a Gaussian distribution? Does that matter?
The next day you walk back up again. You go to the elevation
corresponding to 1.5 sigma. You take a piece of string and on your
map, you trace round the contour level that corresponds to your
current elevation. Question: What is the length of that piece of
string? You walk a bit further (to 2 sigma, say). You again get out
your map and string. Question: Is the length of string longer or
shorter than it was previously? Is your position any more "precise"
than it was earlier? Is that a sensible question? Shouldn't the
relative lengths be the other way round if they have anything to do
with the Gaussian distribution?
I hope that clears things up a bit.
