Automated measurement of Drosophila wingsReportar como inadecuado




Automated measurement of Drosophila wings - Descarga este documento en PDF. Documentación en PDF para descargar gratis. Disponible también para leer online.

BMC Evolutionary Biology

, 3:25

First Online: 11 December 2003Received: 06 September 2003Accepted: 11 December 2003

Abstract

BackgroundMany studies in evolutionary biology and genetics are limited by the rate at which phenotypic information can be acquired. The wings of Drosophila species are a favorable target for automated analysis because of the many interesting questions in evolution and development that can be addressed with them, and because of their simple structure.

ResultsWe have developed an automated image analysis system WINGMACHINE that measures the positions of all the veins and the edges of the wing blade of Drosophilid flies. A video image is obtained with the aid of a simple suction device that immobilizes the wing of a live fly. Low-level processing is used to find the major intersections of the veins. High-level processing then optimizes the fit of an a priori B-spline model of wing shape. WINGMACHINE allows the measurement of 1 wing per minute, including handling, imaging, analysis, and data editing. The repeatabilities of 12 vein intersections averaged 86% in a sample of flies of the same species and sex.

Comparison of 2400 wings of 25 Drosophilid species shows that wing shape is quite conservative within the group, but that almost all taxa are diagnosably different from one another. Wing shape retains some phylogenetic structure, although some species have shapes very different from closely related species. The WINGMACHINE system facilitates artificial selection experiments on complex aspects of wing shape. We selected on an index which is a function of 14 separate measurements of each wing. After 14 generations, we achieved a 15 S.D. difference between up and down-selected treatments.

ConclusionWINGMACHINE enables rapid, highly repeatable measurements of wings in the family Drosophilidae. Our approach to image analysis may be applicable to a variety of biological objects that can be represented as a framework of connected lines.

Electronic supplementary materialThe online version of this article doi:10.1186-1471-2148-3-25 contains supplementary material, which is available to authorized users.

Download fulltext PDF



Autor: David Houle - Jason Mezey - Paul Galpern - Ashley Carter

Fuente: https://link.springer.com/article/10.1186/1471-2148-3-25







Documentos relacionados