| Abstract Detail
The Power of Movement in Plants Miller, Nathan D. [1], Durham, Tessa L. [2], Spalding, Edgar [1]. A Computer Vision Aided Study of Arabidopsis Seedling Root Gravitropism. Darwin carefully studied plant growth and development with the best techniques of his day - glass plates smudged with soot, India ink marks, and his exacting eye. Computer vision, today’s method of choice, has profoundly changed the game by adding precision, objectivity, repeatability, and speed. Electronic images of plants in motion are now captured by high-resolution cameras and the developmental details are algorithmically extracted from the digital movies. Gravitropism, a model stimulus-response system and the subject of this work, lends itself well to this experimental approach. Images of the Arabidopsis seedling root undergoing gravitropism were acquired every 2 min during a 10 h period with 5 µm resolution. Tip angle, rate of root elongation, and root width as a function of time were automatically quantified. Automation enabled a systematic grid of conditions (seedling age, media composition, size of seed) to be explored with many hundreds of trials. Principal component analysis of the tip angle values showed trends in the gravitropic response across the condition grid. K-means clustering separated the population of responses into three classes with distinct time courses that also showed trends across the condition grid. Wavelet analysis was used to calculate the velocity and acceleration of the tip angle for each of the hundreds of trials. Patterns and relationships among the measured parameters resulted in a rich description of the wild-type response and its plasticity. Next, this approach was used to compare the wild-type response with that of a mutant lacking the Glutamate Receptor-like 3.3 molecule. This mutant displays defective ligand-gated Ca2+ fluxes in its root cells but no growth or development phenotype had been detected before the present study found heritable root tip acceleration phenotypes in some but not all conditions. Computer vision promises to enable functional genomics by finding, quantifying, and relating subtle developmental phenotypes.
Log in to add this item to your schedule
1 - University of Wisconsin - Madison, Department of Botany, Madison, WI
2 - University of Wisconsin - Madison, Department of Botany and Cellular and Molecular Biology Program
Keywords:
gravitropism
computer vision.
Presentation Type: Symposium or Colloquium Presentation
Session: SY1
Location: Ballroom 2/Cliff Lodge - Level B
Date: Monday, July 27th, 2009
Time: 9:15 AM
Number: SY1004
Abstract ID:894 |
