Bill's Lecture 6

Question 1

What is the most common method for solving small molecular structures?

Answer 1

Direct Methods

Question 2

How can you solve the structure of larger molecules such as DNA or viruses?

Answer 2

Heavy-Atom Derivatives/Isomorphous Replacement

Question 3

How do heavy atom methods of structure solving work?

Answer 3

Label the macromolecule with heavy atoms.
Different heavy atoms alter the diffracting properties of the samples.
Commonly used metals such as Gold (Au) or Mercury (Hg).
This method requires isomorphous crystals for good results.
Fph1 - Fp = Fh1, Fh1 are the isolated heavy atom vectors.
Allows calculation of a “difference Patterson” using (|Fph1 - Fp|)². Effectively finds area where Argand circles of the molecules overlap.

Question 4

How do direct methods of structure solving work?

Answer 4

Trial-and-error approach.
A probability method of estimating phases. Assume that in the unit cell there is only positive density, that the cell consists of discrete atoms and that phases are
either + or -.
Take three strong reflections: there are eight combinations of + and – phase and for each
combination the phases of many other reflections can be predicted.
Phase sets can be used to calculate maps.
Highly automated using computational methods to interpret the electron density.

Question 5

What is R-factor? and what is it’s formula?

Answer 5

Refinement-factor, used to “refine” the first model since the first model will have errors and is incomplete

R-factor = [(∑||Fo| - |Fc||)/ ∑|Fo| ] x 100%
Fo is the observed structure factor (from the data)
Fc is the calculated structure factor (from the model)
Perfect data and a perfect model would give R-factor = 0%. So if R-factor is decreasing the model is getting better.

Small molecules refined by a minimised term:
∑w|Fo-Fc|², where w=weight linked to the standard deviation of a measurement.