Data analysis

Gerhard Nebe-von-Caron Gerhard.Nebe-von-Caron at unilever.com
Mon Nov 25 08:30:23 EST 1996



          The reference you (and I) look for is of Jim Watson in 
          Cambridge. He has published a book and a nice summary 
          article on Data Analysis, but I can't find it. I am afraid I 
          lent it to someone.

          Gerhard.nebe-von-caron at unilever.com


______________________________ Reply Separator _________________________________
Subject: Data analysis
Author:  kschafer at lucy.tli.org at INTERNET
Date:    23/11/96 01:51


Greetings Flow-ers,
     I have been struggling with a bit of a data
analysis problem and am hoping for a little input (OK,
maybe a lot).  I have been looking at expression of
cell cycle-related proteins in cells in a time course
manner from 0 to 48 hrs after treatment.  The rub is
that the protein expression levels may (and do) either
go up or down compared to the control (0 hr time point)
depending on the length of time after exposure.  What
is the best way to express the change in protein levels
as a number:  using subtraction? or comparing geometric
means? or some other method?  I hope to be able to
statistically analyze replicate treatment time courses
as well.  It is my understanding that subtraction
methods are used primarily for increased levels above
the control (i.e. only shifts to the right measured).
Is there a way to do subtractions that take both into
account?  Are calculations of geometric means
sufficient?  If so, what is the best way to express the
change from the control geometric mean (percent,
absolute [channel number], or other)?  I have explored
Melamed's and Shapiro's texts, but haven't found much
discussion in those references on this particular type
of analysis.
Thanks for your input in advance.

Ken Schafer
Lovelace Inhalation Toxicology Research Institute
Albuquerque, NM
505-845-1126
kschafer at lucy.tli.org



More information about the Cytometry mailing list