10:00a – 12:45a
- need images of 715 beads to align A750 and A647 channels
- Making new 715 bead slide at 1:20,000. If this looks good we can try a 2-color bead slide as well.
- trimming the tape so the epoxy makes glass-to-glass contact all the way around seemed to work well last time for preserving the seal.
- should take conventional overview first at low power — would be nice to have conventional spots to overlay.
- buffer and dye optimization: see notes
analysis of previous 750-647 double stains
- localization number is way down in 647 channel
- Either the probes aren’t working nearly as well
- something went wrong with the buffer (I just changed all the components, new BME aliquot, new COT. Maybe the concentrations and the ratios are actually different).
- plenty of 647 spots are completely undetected as STORM images in 750. The conventional spots aligned very well so I think this must be another issue.
- most of these movies are useless. Should go through and cull the list to save disk space.
- The z astigmatism is still off axis in y. In fact it looks like the astigmatism is non-perpendicular — (on axis in x and off axis in y). This is problematic. Let’s tweak the lens position a little more and see if it can be fixed.
- Need to figure out a good way to work with the images taken in different parts of the quadview.
- We want this to work for display purposes and for image overlay purposes.
- Why don’t we have RunDotFinder modify the .info file for each movie to record the parameter file used to analyze it. Then STORMrender can get the parameter file from the info file and pull out the target ROI.
- now implemented updated RunDotFinder and STORMrender. Needs testing (need new analysis to complete).
- to implement:
- have the option to run silently
- have the option to change mlist name
- changes now implemented. Still need to test.
- changes synced up to alistair branch.
- passage cells
New cell stains:
- F12-P1 + G1-P3
- F12-P1 + F11-P3
- G1-P1 + G2-P3
- Overnight run did not get a great range of PhM values to give a substantial spread of fraction clustered.
- Rerunning with higher variance (3) and more rounds (30 instead of 20).
- maybe should optimize to get a larger range of response for PhM concentration effects first…
- rerunning PhM concentration effect simulations with N=500 instead of N=300.
- double check probe uniqueness in latest data set
- probes not unique when Tm is lower than Cx temp.
- checking if probes are indeed unique except for same gene matches when Tm is greater than Cx temp.