User f3aa6e9bea
10-02-2014 19:04:21
Hi
together,
during my
master´s thesis I have extensively used your great geometry plugin in order to
calculate minimal and maximal projection radii. In my thesis I want to explain
shortly what happens in the plugin to calculate the radii. If I had understood
prior explanations in this board correctly it does the following:
1. It takes
my 2D sketch and calculates the lowest energy conformer using MMFF94
2. Discrete
orientations of the molecule are projected on a plane.
3. An
algorithm finds the projection radius extrema
4. The
molecular orientation for the projection is fine tuned by a numerical optimizer
(projection optimization)
Have I
understood the principle correctly?
One other
question: When is it necessary to deselect projection optimization? Only if
calculation time is otherwise too long or are there other reasons e.g. special
molecules where I should not run the optimizer?
Regards,
X4U32
ChemAxon 1b9e90b2e7
21-02-2014 23:12:06
Hi Henrik,
thank you for your post .
"Have I
understood the principle correctly?"
Yes, that is correct. This is the outline of the minimal-maximal projection surface area method.
"When is it necessary to deselect projection optimization?"
We apply a function like proj.surf.area(orientation of the molecule). As you have correctly pointed out, this function is sampled, and the orientation for the extrema is returned. Optionally, this orientation can be further optimized by a numerical optimizer. In order to be optimizable, this function needs to be modified. As a result, the non-optimized and optimized projected surface area values are not comparable. In other words, if you have a set of molecules to run the proj. surface calculations on, use the same settings.
According to some small scale benchmarking, the initial sampling is dense enough, which means no strict need for the "projection optimizaion" to be enabled.
It would be nice, if you wrote couple of words about the bigger picture, how these values are used in your project.
Cheers,
Adrian