10:00 A – 8:00P, 9:30P-10:30P
- In situs day 1
- Flip fly cages, 10:30A
- Attempted sodium borohydride treatment of half of samples. Left sealed on shaker — vials exploded, rocket off of shacker from underneath tin foil cover half way across desk. Samples lost.
- Continuing with remaining embryos.
- Stain MP09 with y-int dig. y-full bio.
- Stain wt with E(Pc)-dig, en-flu, tou-bio, see if we can recapitulate earlier staining.
- Fix more MP09 embryos, stain with Bcd 2 hrs.
- stain PcGs. Co-stain later with K27 and K4 as comparison and control. Try to get early embryo time series data (embryos are 45min – 4 hrs old).
- Working on ACS app
Coding Drift Correction
- working on drift correction. Fixed data output bug, just track position from fread using count output and advance to next frame with fseek.
- Attempting to drift correct seemingly blurred 647 channel from Hb data. Channel only looks blurred using multi-emitter fitting (perhaps single emitter not enough localizations to tell, but this is still a bit weird). More weird, even at 750 frames a batch the individual sub images still appear smeared in x, but the total drift is actually small.
- single emitter fitting not dense enough to provide good fit — should have a parameter for this, if correlation is not good enough integrate longer or give up fit.
- Discuss drift correction with Hazen: Should fit 2D Gaussian to 2D cross-correlation peak and use the max of that, rather than the max pixel.
- Hazen has python script to do drift correction from molecule list *_list.bin files. Should be able to call this from matlab.
Image analysis
- Attempting to run GPU analysis from Tuck on Alistair3b hard-disk connected to Monet.
- Setup to run Insight on Tuck from data drive connected to Monet. Runs a bit slow by executes fine.