Wednesday 04/23/14

8:30a – 10:45p

Career planning

from Jeff: K99 next application deadline is June 12th. After that it’s October 12th (see NLM page, NIGMS not as clearly laid out).
recommended to apply with time to submit resubmission ~8 months after original application. Application and any resubmissions must be received before the applicant starts the 4th year of post-doctoral training. This un-necessarily ambiguous terminology is actually strictly interpreted as 4 years from the date on your PhD, it doesn’t seem to matter what you’ve been doing with the exception of clinical training or parental leave. See FAQ
This means to have a hope of being able to submit a revision in response to reviewer feedback, I need to apply in June.

copying data off temp-drive on STORM2
keeps on failing / interrupting due to system restarts in attempt to get lasers and fluid controls to both communicate properly with computer.
cancelled data copying. With the temp-drive now recording data and the target drive analyzing data copying more data from temp to target is impossibly slow and it good Windows fashion makes the analysis impossibly slow as well.
CopyOnArrival functions in matlab-functions don’t copy folders, just files. Would be good to fix this.

calibrating DaoSTORM parameters for second round L3E01, L3E07, L2F7F6, L2F7F6 pt 2
running DaoSTORM fitting on second round L3E01, L3E07, L2F7F6, L2F7F6 pt 2
haven’t been good about getting regular chromatic and z-calibration beads for recent data sets. Need to get better.
running analysis of STORM2 beads from 03/28 data set to apply to L3C01-C08 data for the runs taken on STORM2.

current CalcChromeWarp is a little bulky, doesn’t do dual view.
rewriting CalcChromeWarp from scratch
Using region finding instead of splitdax to speed things up a bit. This has a few new challenges:
- insightM overwrites filenames, since filenames have to match daxnames and all daxnames are the same
- insightM gives pop-up warnings (InsightM crashed) when movies are blank (e.g. no 488 data). This is annoying.
- solve this by switching to single iteration DaoSTORM
- ROI is going to have offsets — beads indexed to their positions in the original file. Will need correct these offsets.
- fortunately I save mlist_pars, should be able to correct.
substantial changes to RunDotFinder to be able to pass alternate binnames. Maybe want to switch back to development branch for regular analysis until we are ready to take this further.