Identification of molecular markers associated with sweet potato resistance to sweet potato virus disease in Kenya
Sweet potato virus disease (SPVD), a result of the co-infection of whitefly transmitted Sweet potato chlorotic stunt virus (genus Crinivirus, family Closteroviridae) and the aphid transmitted Sweet potato feathery mottle virus (genus Potyvirus, family Potyviridae), is the most destructive disease of...
Gespeichert in:
Veröffentlicht in: | Euphytica 2008-03, Vol.160 (1), p.15-24 |
---|---|
Hauptverfasser: | , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | Sweet potato virus disease (SPVD), a result of the co-infection of whitefly transmitted Sweet potato chlorotic stunt virus (genus Crinivirus, family Closteroviridae) and the aphid transmitted Sweet potato feathery mottle virus (genus Potyvirus, family Potyviridae), is the most destructive disease of sweet potato in East Africa. A study was conducted to establish if genotypes identified as resistant or susceptible to SPVD in Kenya could be distinguished using molecular markers. A total of 47 unrelated sweet potato genotypes were selected from germplasm collections and classified into two phenotypic groups as resistant or susceptible to SPVD. Genotype selection was based on disease severity or days to symptom development in plants following graft inoculation. Amplified fragment length polymorphism (AFLP) marker profiles were generated for each individual and used in association studies to identify markers suitable for classifying the two pre-defined phenotypic groups. Analysis of molecular variance showed significant (P < 0.002) variation between the two groups using 206 polymorphic AFLP markers. Discriminant analysis and logistic regression statistical methods were used to select informative markers, and to develop models that would classify the two phenotypic groups. A training set of 30 genotypes consisting of 15 resistant and 15 susceptible were used to develop classification models. The remaining 17 genotypes were used as a test set. Four markers, which gave 100% correct classification of the training set and 94% correct classification of the test set, were selected by both statistical methods. |
---|---|
ISSN: | 0014-2336 1573-5060 |
DOI: | 10.1007/s10681-007-9495-2 |