9:00a – 5:00p
Goals
- set up STORM imaging for PhWtPh.
- improved drift correction
- write quick script to average bead dax movies (and split them out?). Test on the G1-G1 0004 data?
- modify feducialDriftCorrection function to be able to handle time averaged dax movies and spread them out a bit.
- plot auto-correlation in non-time averaged movies vs time averaged ones.
- plot inter-dot correlation in non-time averaged movies vs time averaged ones.
- Adapt ChromatinCropper for multicolor data.
Buffer exchange chambers
- smaller channels definitely work better.
- need to keep buffer supply at equal height as chamber — otherwise capillarly action plus the potential difference due to the elevation drives positive pressure into the chamber, breaking the seal. Need to maintain negative pressure in the chamber.
Ph project
observations on transfection rate in PhM (upper right panel is endogenous Ph). There is strong correlation of the bright cells in the endogenous channel and the bright cells in the Ph-flag channel.
STORM imaging Ph
- buffer conditions for Ph-STORM: 5 uL MEA + 5uL COT + 10 uL GLOX + 100 uL 50% glucose + 400 uL Im buffer (200 mM Tris)
- Cy7 is sufficiently bright with this MEA:COT ratio. Compare with 1 uL MEA for 2 color F12-F11 and F12-G1 images and 2 uL MEA for G1-G2 images.
- Setting up STORM of new Ph-Wt-Ph. sparsely distributed much brighter cells.
- Clustering much more evident it seems.
- power settings: 2350 mA of 750, 400 mW of 647.
- 205K frames of 750 switching, mostly with no 405 (could have gone many more, but that’s where it starts slowing down)
- 105K frames of 647 switching (also where it just starts slowing down).
- some cells do have clear nuclear expression at much, much lower levels than the few bright one that jump out at you.
- off and running at 3pm (estimated time for complete run, 48 hours. most of its at 30hz, so we won’t get close to that).
Ph Psc data review
- data set paralleling the Pc data doesn’t look too bad qualitatively. Never analyzed the STORM.
- Launched analysis for Ph-Psc data from 2013-09-26.
- started analysis of PhMPh dataset. This pretty much saturates my connection to the data disks so the rest will have to wait to get launched.
More data shuffling
- (definetely going to striped or another RAID array next).
- 2013-12-20_G1-2clr moved from ProBox6 to ProBox4 (need to delete orig when done)
coding matlab-storm
- updated color options
- made
ColorByFrame.mits own function. This one is better than the version used by ChromatinCropper, I should update that. - STORMfinder not obeying new run silently / run in background options.
- fixed — issue with run in background vs run in matlab, it matters how you ask the question.
- damn it, out of of disk space on ProBox4 to print the new daxfile. Hence a pile of issues in trying to analyze it.