MOHCA analysis
Rhiju Das
rhiju@u.washington.edu
Summary of code developed up until Feb 2006.
Posted on-line on Sep. 22, 2007

There are several challenges in analyzing MOHCA gels.
The main one is signal-to-noise -- is a spot reproducible?
A second one is whether the spot can be attributed to
background cleavage.
The only way I've figured out to answer these questions is
to run several gels. Then I try put them all in the same coordinate
system -- do the hits on one gel show up in similar
locations on all the gels?

There are three overall alignment steps:

(1) Run a really nice 2D gel with a 5'labeled T1 grid. This produces
marks that correspond to G's in both the x and y directions.
The rectified version of this gel becomes the overall standard for
all other experiments with 5'labeled MOHCA expts.

    The alignment steps for this gel are in: 
    rd149a_GTPalphaS_top.m

    You could see what happened when I ran these commands if you load the following workspace:
    standard.mat

(2) Run a nice 2D gel with a T1 grid overlaid with a MOHCA pattern. 
In this case I ran A-bomb, 5prime-labeled samples for P4-P6 in the
unfolded state. This lets me see very accurately where the MOHCA hits 
are, relative to the T1 grid. It would be nice to run T1 grids overlaid
with everything, but they tend to cover up some MOHCa hits; in the future
 it would be great to run them in a different color, if fluorescent
labeling becomes useable. 

    The alignment steps for this gel are in: 
    A5prime_standard_rd91g_2d.m

    You could see what happened when I ran these commands if you load the following workspace:
    standard_A5prime.mat
    

(3) Finally run a MOHCA gel (or actually several repeats).
This gel then gets rectified as best as possible, and an approximate
coordinate system is set up. The coordinate system is then refined by
a final alignment to the gel #2 above. The last step is then to define potential hits; a routine then estimates a background based on cleavage patterns in
adjacent lanes, and you get a confidence number (signal over noise) for the hit.

    The alignment steps for this gel are in: 
    rd150d_p4p6_ABOMB_5prime_Unfolded_MKtopB.m

    You could see what happened when I ran these commands if you load the following workspace:
    saverd150d_p4p6_ABOMB_5prime_Unfolded_MKtopB.mat
    
