Monday 04/01/13

10:00 A – 7:00P, 9:00P-10:00P

General

Insight 1: enhancer switching between obstructed and unobstructed case is the bursting model with different tx rates from ‘off’ and ‘on’ promoter states. (and non-exponential waiting times).
No decay -> law of large numbers is just going to send the variance to zero.
Fast switching also just going to get the law of large numbers average of the obstructed and unobstructed states
Insight 2: can’t just look at raw CoV as a function of A — the effect of low TF on noise washes out the effect of the promoter architecture. Should look at fold change for each value of A.
changed ModelMoments to return functions mu_vs_A and Cov_vs_A

Running DaoSTORM Analysis on Monet of Ph-FLAG (11 movies).
Analysis competing with copy (previously copied from Temp1 to Temp2, not to Alistair10).

planning experiments for oligo-paint probes
GREEN and YELLOW probe labels SWITCHED in my order form Brian’s library. Must check and correct
Ordered primers

try less overhang on tertiary (2 bp? no bp? get that 405 closer).
make secondary longer — 32 bp homology + space in-between, allows homology of PPT primary which may lose some bases.

MATLAB crashed, 5PM.
won’t re-launch.
closed additional version of matlab running STORM analysis
Matlab still won’t re-launch
Attempting to remote desktop to Tuck or Cajal to use their Matlab
Remote Desktop won’t connect.
Saving processes, aborting all analysis, restarting computer.
Restarted copying data
Restarted analysis. Picked up at frame ~1000, I thought it was closer to 30,000, but maybe that was the molecule number. Vaguely worried it might be double counting / not resuming analysis correctly….
Matlab goes java lang error beserk (both versions, running analytical manipulation of matrices and running spotfinding). Closed both