|
Development of a predictive model for northern corn leaf blight (NCLB) resistance in maize through SSR markers association and validation |
|---|---|
| รหัสดีโอไอ | |
| Creator | Theerawut Wongwarat |
| Title | Development of a predictive model for northern corn leaf blight (NCLB) resistance in maize through SSR markers association and validation |
| Contributor | Panuwat Sinlapasakkajohn, Suwara Wuthiaumphon, Chalong Kerdsri, Chaowanart Phruetthithep |
| Publisher | Mahasarakham University |
| Publication Year | 2569 |
| Journal Title | Food Agricultural Sciences and Technology (FAST) |
| Journal Vol. | 12 |
| Journal No. | 2 |
| Page no. | 94-106 |
| Keyword | Northern corn leaf blight, single marker analysis, SSR marker, maize resistance, marker-assisted selection |
| URL Website | https://ph02.tci-thaijo.org/index.php/stej/index |
| Website title | Food Agricultural Sciences and Technology (FAST) |
| ISSN | 2822-1001 (Print), 2822-101X (Online) |
| Abstract | Northern corn leaf blight (NCLB), caused by fungal pathogen Exserohilum turcicum, is a major foliar disease of maize, particularly under cool and humid conditions during the dry season (November-February) as observed at the Chai Nat Field Crops Research Center. Severe infections during the reproductive stage can significantly reduce grain yield. Although chemical fungicides are available, their environmental impact makes host plant resistance a sustainable solution. This study aimed to assess the association and predictive accuracy of simple sequence repeat (SSR) markers for NCLB resistance, to support efficient selection of resistant genotypes in maize breeding programs. Three SSR markers, bnlg198, umc2038 and umc2210, were evaluated for their association with NCLB resistance in 280 sweet corn genotypes derived from two populations, namely CH66C1 and HX75C1. Chi-square test was initially used to perform single marker analysis in order to assess the association between each SSR marker and resistance to NCLB. Markers that showed significant associations were subsequently used to construct prediction models using regression analysis. Model performance was then evaluated using Adjusted R2, F-value and receiver operating characteristic (ROC) curve analysis. Five-fold cross-validation was employed to estimate predictive reliability. All three markers, bnlg198, umc2038, and umc2210, were significantly associated with NCLB resistance. The model, including umc2038 and bnlg198, showed a positive correlation coefficient of 0.6, an adjusted R2 of 0.57 and the highest F-value (36.4), indicating that these markers explained 57% of phenotypic variation. The model's predictive accuracy was classification as moderate to high with an area under the ROC (AUC) curve of 0.78. The best threshold identified was 0.51, yielding a true positive rate (TPR) of 0.88, indicating that the model could correctly classify 88% of resistant lines. |
