Tuesday, January 31, 2006

Fun Times

What are you doing right now? I bet it is not as exciting as what I'm doing. If you want your kids to become scientists, let them have a looks at my glamorous day:

Ready, Set, Go:

*open green channel
open red channel
convert green channel to grayscale
save picture in new folder
switch programs
open grayscale channel 1
set particle size
switch programs
grid out targets on channel 2
switch program
circle target 1 on inverted channel 1
analyze particles
copy measurements
switch programs
paste into spreadsheet
move over one column
switch programs
**circle next target on inverted channel 1
switch programs
make sure the target you circled is correct on grid created over channel 2
switch programs
analyze particles
copy measurements
switch programs
copy into spreadsheet
***move over one column
repeat from ** to *** until all targets on channel 1 are analyzed
repeat procedure from * to *** until all 150 pictures are analyzed

It's like knitting...sort of.

Why don't I used software that automates all of this, you are wondering? I would LOVE to say software like that doesn't exist. It does, however, exist. It's just not free. Free is my spending limit. Being that Kelly Ripa makes more in one second of working (I learned this on VH1's Fabulous Life of Kelly Ripa) than I do in one hour, the cost of my time isn't worth gifts of time- and energy-saving software. I know I've complained about this before...

I should have been a dancer.

Not the exotic kind, though. I'm sure the spin around the pole, take off top, spin around the pole routine gets just as tedious as data analysis.


At 11:09 AM, Anonymous angiebean said...

Hey. I sympathize with you. I am doing confocal images and trying also to animate them in z-sections. My PI doesn't know anything about confocal microscopy and keeps on saying I need to prove that the two channels are actually colocalized and not just bleed through (ugh!).

Any suggestions on how to use photoshop to prove my point?
Please email me at:

Thanks for the help. Really like the blog!

At 11:31 AM, Blogger she falters to rise said...

Z-stacks--love 'em. I am actually getting a new computer because my superpower laptop can't handle my 3-D animations. The Dell people couldn't believe it, but alas, it is true.

In order to prove that it's not bleedthrough, you need to capture the two channels sequentially instead of in tandem. I had a lot of problems with my alexa-fluor 488 showing up on my 568 images so I now capture each channel individually as standard practice. Photoshop can not and should not be used for "data" analysis--it should only be used for making things pretty as long as you don't change the information in the process.

You could use metamorph and figure out the %co-localization you have per cell. If you don't see 100% co-localization with your brightest signals, then you could start to claim that it isn't bleedthrough.

The best way to "prove" co-localization is to do both co-focal, dual channel overlays and also FRET analysis because there is some error associated with co-focal overlapping. FRET can be a pain though and you don't see a lot of people using it yet.

At 11:05 AM, Blogger sue said...

Okay, my head is spinning now. Stop with the geek talk, 'kay?


Post a Comment

<< Home