11:30 A – 1P, 1:45P – 12:45 A
STORM
- Overnight run on STORM3 crashed at around 34 of 85 movies — ‘memory error’.
- send crash report to Hazen.
- copying data form Alistair9 to Alistair8.
- Imaging BX-C embryos with H3K27me3. One nice older embryo with clear AP distinction. Will try to get multiple images of this guy tonight. Hopefully all the STORM channels cooperate.
- 561 and 488 channels have high background. 488 impossible to bleach out, always get full nuclei glowing with 405. Can’t break down to individual nuclei (I wonder if any of the DAPI tail is still staying around?)
- For next time: Try thinner sections, 400 or 500 nm max. NBrH4 treat coverglass prior to staining.
Coding
- Starting merge of Matlab-STORM (my current, active development folder), into matlab-storm (my shared project folder with Jeff).
- created new projects in matlab-storm Beta: STORMfinderBeta and STORMrenderBeta. These are the “in development” versions of the same files with the non-beta names. The Beta should allow me to keep everything in the matlab path without confusing matlab. The non-beta versions are the current stable releases. This should allow me to do some more development without compromising stability or having half-active functions.
- Fixed bug in DaoSTORM simplified install — needed to change filepaths to drift correction and molecule linking scripts, since we moved mufit_anaysis.py down into the base DaoSTORM folder, the ‘..’ look back a directory creates errors.
- Test ran DaoSTORM on Monet, working great.
Issues with Github creating merge conflicts
- Solution: Git pull first, that will get you the latest stuff from the online Repo and create a merge conflict
- Open the indicated merge-conflicted file in an editor. Find the text bracketed by <<<<<<< ========= >>>>>>> and change the code to what you want (your version, their version, or some mix).
- Add the file git add, and git commit.
- Now git Push will see you’ve resolved the conflict as you’ve done a recent pull and have files that have no merge conflict flags.
- More details here:
Code planning with Jeff: meeting notes
- matlab-storm folder should be essentially stand-alone (not require additional software)
- GUI development: GUIs should support multiple instances
- all global variables used by GUIs become cell arrays, indexed by instance number
- Each instance knows its number so it knows where in the array to grab.
- Close functions may need to be overwritten to clear the appropriate fields
Fly Work
- collect Mat. alpha tub 2x. virgins (AM)
- collect Mat alpha tub VP16 on 3 virgins, crossed to yw; Sp/CyO; TM2/TM6
- collect Mat. alpha tub 2x. virgins (PM)
- collect Mat alpha tub VP16 on 3 virgins, crossed to yw; Sp/CyO; TM2/TM6 (PM)
Data Analysis
- use new version of STORMfinderBeta to select DaoSTORM parameters. Fixed bug in zfit. Updated default DaoSTORM parameters.
- Started running first batch of BX-C imaging in embryos through DaoSTORM.