Message Boards Message Boards

1 Reply
1 Total Likes
View groups...
Share this post:

Ancova analysis on two linear regressions (from microscopy data)

Posted 11 years ago
Hello everybody,

I would like to kindly ask for your help.
the results of the microscopy analysis I perform on a specific specimen are two images taken with different wavelengths (405 and 488).

It is possible to plot the intensities of a specific pixel at 405,x and 488,y (in a scatter plot for example) and to fit a linear regression line.
Changes in the regression slopes are observed depending on the treatments performed to the specimen, between condition A and B.
I would like to determine if two regressions are statiscally significant. I believe I need to perform an ANCOVA for that. However, I still wasn't able to compute it in mathematica specially because I have problems describing the nominal variable.

Could you please give me some tips?
Thank you very much for your time,

This is my procedure so far:
Loading 4 images: A405, A488, B405, B488
PixA405 = Flatten[ImageData[A405, "Bit16"]];
PixA488 = Flatten[ImageData[A488, "Bit16"]];
PixB405 = Flatten[ImageData[B405, "Bit16"]];
PixB488 = Flatten[ImageData[B488, "Bit16"]];
TabA405A488 = Table[{PixA405[], PixA488[]}, {n, 1, PixNo}];
TabB405B488 = Table[{PixB405[], PixB488[]}, {n, 1, PixNo}];
LinearFit1 = LinearModelFit[TabA405A488, {x}, x]
LinearFit2 = LinearModelFit[TabB405B488, {x}, x]
POSTED BY: Ana Gomes Barata
I am not super strong in statistic and I am not sure if I understood your problem correctly.
However you don't need LinearModelFit to fit a line between two points,

sA = (Flatten[ImageData[A488, "Bit16"]] - Flatten[ImageData[A405, "Bit16"]])/(488-405)

sB = (Flatten[ImageData[B488, "Bit16"]] - Flatten[ImageData[B405, "Bit16"]])/(488-405)
already contain all the slopes.

What kind of relation do you expect between sA and sB (namely the effect of the treatments)?
Reply to this discussion
Community posts can be styled and formatted using the Markdown syntax.
Reply Preview
or Discard

Group Abstract Group Abstract