9:30 A – 7:30P (lab) 8:30P – 2:40A (remote)
Coding
- troubleshooting issue with Insight’s Z-fits not matching mine.
- Try flipping wx and wy?
- try flipping zc = zlevel – zstage
- Maybe the issue is in the write inifile / which inifile temp inifile is getting called through STORMfinderBeta, AutoZCal, and CompZcal.
- This seems to fix things
- Calibration from STORMfinderBeta with Insight now looks good.
- Testing DaoSTORM
- Need to ensure z-positions are monotonic for interpolation (no idea why this didn’t object when running insight, its the same dax file same .off file with stage offset info). No fixed. Interpolation is still very fast
- Simulated annealing 10% shifts is a good start, 30% can be too much.
Antibody labeling
- UV-vis best sensitivity 0.1 – 1 absorbance range, over 2 not very accurate — dilute the sample, not enough light is reaching the detector.
- 2 rounds, 2nd round gives reasonable 750 labeling (with 9 uL of 2 ug/uL)
Fly Work
- Collect virgin 2x mat alpha tubs. some dark eyed non-Virgin flies found on pad. Stock contamination issue?
- Scm and Ph mutants need flipping, but we are out of bottles. Try later.
- Too much snow, skipped evening virgin collection.
Meeting prep for discussion with Xiaowei.
- preparing slides to update
- presentation can be viewed from chromebook remote desktop, just remember to unplug monitor 2 prior to viewing.
- see meeting notes. Experimenting with loading presentation as a google doc and embedding into post. Not working at present.
- Interesting patterns in Scm data, see post.
STORM analysis
- Ran z-bead calibration
- Problem with DaoSTORM bead calibration is on STORM2 bead movies it likes to split the out-of-frame beads into 4 separate beads with smaller widths.
- In general probably need a more restricted z-range with DaoSTORM. Should also bias image acquistion towards the z-direction that does less splitting.
- Possible issue with different definitions of above and below plane in DaoSTORM too. Though I didn’t have this problem with STORM3… maybe it’s a system stage encoding difference? Definitely need to track this error down.
- Stage should go back the way I had it for both STORM2 and STORM3 data (need to check DaoSTORM vs Insight still). Some weirdness yet in +/- z
- annealing entirely unnecessary — just use crossing point of wx and wy, this should ID when any bead passes through z = zero.
- Also annealing approach has the bad idea of uncoupling these two, and the left right random jigger leaves a random error in the recoupling. Plus its much slower than the new method.
STORM3 troubleshooting crash: TO Do:
- Rerun same script (with added positions)
- Shorten all STORM movies to 5 frames
- see if we can repeat the crash
- if so, check system memory from cmd after crash
- if Hal freezes, launch crystal AOTF controller (labview based), which will steal control of AOTF and allow hal to exit.