Causal Graphs
These are a mechanism for formally describing possible failure cases in
DNA. At the moment we have just looked at the screening process,
identifying a couple of things which could have gone wrong, and how we
would see these things.
This process does not offer any solutions, simply a mechanism for
finding and identifying problems.
Here is an example:

This is a very simple example, which contains two mechanisms for
detecting faults (detectors) and three possible faults. For an example,
I will talk through the 'icy crystal' situation.
An icy crystal (i.e. with 'ice' forming on the surface) will cause
curcular diffraction patterns on the image - ice rings. The symptom of an icy crystal is
therefore ice rings on the diffraction pattern. These ice rings can
exhibit their presence in three possible ways:
- Mosflm 'fails' in the indexing - either giving no solutions or
- Mosflm 'fails' by producing a silly solution (i.e. 4 x 5 x 200 A
cell) or
- The DiffractionImage module screens the image and finds ice rings
These indicators should
therefore be able to clearly say that this problem occurred. These make
use of two detectors, Mosflm
and the DiffractionImage module. Without the DiffractionImage module it
would not be easy to distinguish between this situation and others.
Clearly, in a non automated system there is a further detector, the
user, and he or she may spot these possible indicators of failure.