Unable to connect to database - 11:22:02 Unable to connect to database - 11:22:02 SQL Statement is null or not a SELECT - 11:22:02 SQL Statement is null or not a DELETE - 11:22:02 Botany & Plant Biology 2007 - Abstract Search
Unable to connect to database - 11:22:02 Unable to connect to database - 11:22:02 SQL Statement is null or not a SELECT - 11:22:02

Abstract Detail


Seed Biology

Moore, Candace Randall [1], Miller, Nathan D. [2], Vaughn, Laura M. [3], Spalding, Edgar [4], Masson, Patrick H. [3].

A Computer Vision Based QTL Analysis of Arabidopsis Seed Morphology and Color.

Seed size has been associated with plant fitness and therefore its genetic determinants should be identified. Quantitative trait locus (QTL) mapping in Arabidopsis has identified chromosomal regions that control seed traits such as length (1), but the time required to measure many seeds has inhibited advancement. This constraint can be alleviated using computer vision tools which harness computational algorithms to extract meaningful information from electronic images. A high-throughput method of analyzing seed shape and color based on computer vision was developed and applied to seeds produced by Cvi x Ler2 recombinant inbred lines for the purpose of QTL mapping. The process of quantifying seed shape and color begins with the digital imaging of groups of approximately 1000 seeds with a standard flatbed scanner at a resolution of 127 pixels per mm. A computer algorithm written in the Matlab language isolates the seed image from the background. This produces a set of objects that are subjected to morphometric analyses resulting in quantification of seed area, length, width, and color (pixel intensity in each of the red, green, and blue channels). Analysis of the results with QTL Cartographer identified seven loci that contribute to seed size (chromosome 1 contained 3 loci, chromosome 5 contained two loci, and chromosomes 3 and 4 each contained one locus). These results agree well with a previous study (1) and include one novel locus. These data highlight how computer vision can aid in measuring large numbers of samples for QTL analysis. This method could prove useful for quickly measuring spatial and spectral properties of seeds and other plant structures in many species.
1. Alonso-Blanco C, Blankestijn-de Vries H, Hanhart CJ, Koornneef M (1999) PNAS 96:4710-4717


Log in to add this item to your schedule

1 - UW-Madison, Department of Botany, B133 Birge Hall, 430 Lincoln Dr., Madison, WI, 53706, USA
2 - UW-Madison, Departments of Botany and Biomedical Engineering
3 - UW-Madison, Department of Genetics
4 - UW-Madison, Department of Botany

Keywords:
QTL
computer vision
Seed.

Presentation Type: Plant Biology Abstract
Session: P
Location: Exhibit Hall (Northeast, Southwest & Southeast)/Hilton
Date: Sunday, July 8th, 2007
Time: 8:00 AM
Number: P29025
Abstract ID:2078


Copyright 2000-2007, Botanical Society of America. All rights