9:30 A – 7:30P STORM
* Finish O/N run by 10A.
* Run Z-calibration
* Copy and delete data! Remove external hard-drive
* Image BX-C + H3K27me3-488 slide (at 4C in PBT). maybe we should store these in 2x SSCT? STORM analysis
- Running InsightM using new STORMfinder interface on 3 original BX-C clusters run.
- 3D HP-1 clusters (dense images) and BX-C labeled complex, (zoomed in small images).
- The z-calibration appears to be off — dots span the whole range up and down. This same z-curve on my bead data does very poor.
- Even worse at z = 0
-
Troubleshooting ‘AnalyzeAllDax’ command for method ‘GPUmultifit’
- WriteMoleculeList needs all fields populated. With this fix, the thing now works.
- Troubleshooting suddenly slow processing speed of insightM on large data sets:
- fclose in WriteInfoFile takes 10 seconds
- fclose(ftemp) in write dax file takes 10 seconds
- InsightM takes 77 seconds to find molecules in a single frame
- –> problem of working across network. Especially when folder is already being accessed to copy files from STORM2 computer. Runs fast locally.
-
Updating STORMrender
- Change over to new mlist format (maybe just write a converter as a stopgap?)
- changed plot_STORM and plot_STORM_zcolor to new mlist format.
- All single color functions now use new mlist format
- load drag and drop bin files (into channel of choice?)
- drag and drop bin files now quick load into STORMrender in channel 1 if users selects to use STORMrender. Otherwise a molecule list is read into memory.
- Quick load button now tries to read the global ‘binfile’ (generally defined by drag-and-drop). Fly work
- Change over to new mlist format (maybe just write a converter as a stopgap?)
-
get flies from mailroom
-
collect virgins: sim[D] Espl[D], Communication
-
Accept review request (due Feb 7th)
- Stas proposes I write a short BJ letter on a probablistic model of shadow enhancers. I promised to work this out and send him a draft.
- Reply to Banff team about contacting participants