image analysis - for courses and Discovery Sce

classic Classic list List threaded Threaded
1 message Options
Anya Salih Anya Salih
Reply | Threaded
Open this post in threaded view
|

image analysis - for courses and Discovery Sce

Date:    Fri, 19 Jun 2009 10:49:27 +0200
From:    Daniel James White <[hidden email]>
Subject: questions about FRET - colcoalization - FLIM

Hi Ramana,

regarding colocalization,
you might like to use the relatively robust methods of Manders and
Costes for pixel intensity correlation over space.

you can do it with
Fiji (is just imageJ - batteries included) which has working
colocalisation threshold and Test plugins
http://pacific.mpi-cbg.de

or
BioImageXD
http://www.bioimagexd.net

both for free!

mopst commercial packages dont impement these methods fully.
Auto  threshold is very important,m as is the statistical test to see
iof the correlation you see is better then random.

See my previous posts on this subject... there is a lot to get right.
Have a look here for a practical guide
http://pacific.mpi-cbg.de/wiki/index.php/ColocalizationAnalysis

and here for theory
https://info.med.tu-dresden.de/MTZimaging/images/2/25/White_QuantitativeColocAnalysis0408a.pdf

and here for quantitative  data collection ideas
https://info.med.tu-dresden.de/MTZimaging/images/6/63/White_QuantitativeColocAnalysis0408b.pdf



By the way..... the most reliable way to do FRET is to actually to do
FLIM, fluorescence life time imaging.
It gives the information you want, but gets around many of the
technical problems of traditional FRET that make results unreliable.


cheers

Dan