Hi There,
Wonder if anyone can help ?
I'm working on a program for the rapid identification of Clinical
Mycobacteria.
All fine and well so far, works reasonably well. But I'm looking to
extend the capabilities by incorporating the facility to suggest which
tests to do next that will yield the most discrimination, e.g
You have an identification based on, say, Growth at 25C, Lebek Niveau,
Cat 45C and possibly a BST digest profile. The program returns three
organisms with
an identical match score. Although the program currently displays the
%positives for
a variety (35) of biochemical/enzymatic tests (and from that it is
fairly easy to visually make some discriminative decisions), I would
like to try and automate the process.
Wondered if anybody knew of any off-the-shelf algorithms to help before
I start
developing my own , or had any idea where to start looking ?
Please reply direct to me at the emcphers at csd.abdn.ac.uk address as well
as
posting to the group.
Many Thanks,
Ewen
---
