Dataset Information

Deep Learning Diagnostics of Gray Leaf Spot in Maize under Mixed Disease Field Conditions.

ABSTRACT: Maize yields worldwide are limited by foliar diseases that could be fungal, oomycete, bacterial, or viral in origin. Correct disease identification is critical for farmers to apply the correct control measures, such as fungicide sprays. Deep learning has the potential for automated disease classification from images of leaf symptoms. We aimed to develop a classifier to identify gray leaf spot (GLS) disease of maize in field images where mixed diseases were present (18,656 images after augmentation). In this study, we compare deep learning models trained on mixed disease field images with and without background subtraction. Performance was compared with models trained on PlantVillage images with single diseases and uniform backgrounds. First, we developed a modified VGG16 network referred to as "GLS_net" to perform binary classification of GLS, which achieved a 73.4% accuracy. Second, we used MaskRCNN to dynamically segment leaves from backgrounds in combination with GLS_net to identify GLS, resulting in a 72.6% accuracy. Models trained on PlantVillage images were 94.1% accurate at GLS classification with the PlantVillage testing set but performed poorly with the field image dataset (55.1% accuracy). In contrast, the GLS_net model was 78% accurate on the PlantVillage testing set. We conclude that deep learning models trained with realistic mixed disease field data obtain superior degrees of generalizability and external validity when compared to models trained using idealized datasets.

SUBMITTER: Craze HA

PROVIDER: S-EPMC9330607 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BACKGROUND: Gray leaf spot (GLS) is a globally important foliar disease of maize. Cercospora zeina, one of the two fungal species that cause the disease, is prevalent in southern Africa, China, Brazil and the eastern corn belt of the USA. Identification of QTL for GLS resistance in subtropical germplasm is important to support breeding programmes in developing countries where C.?zeina limits production of this staple food crop. RESULTS: A maize RIL population (F7:S6) from a cross between CML444 and SC Malawi was field-tested under GLS disease pressure at five field sites over three seasons in KwaZulu-Natal, South Africa. Thirty QTL identified from eleven field trials (environments) were consolidated to seven QTL for GLS resistance based on their expression in at least two environments and location in the same core maize bins. Four GLS resistance alleles were derived from the more resistant parent CML444 (bin 1.10, 4.08, 9.04/9.05, 10.06/10.07), whereas the remainder were from SC Malawi (bin 6.06/6.07, 7.02/7.03, 9.06). QTLs in bin 4.08 and bin 6.06/6.07 were also detected as joint QTLs, each explained more than 11% of the phenotypic variation, and were identified in four and seven environments, respectively. Common markers were used to allocate GLS QTL from eleven previous studies to bins on the IBM2005 map, and GLS QTL "hotspots" were noted. Bin 4.08 and 7.02/7.03 GLS QTL from this study overlapped with hotspots, whereas the bin 6.06/6.07 and bin 9.06 QTLs appeared to be unique. QTL for flowering time (bin 1.07, 4.09) in this population did not correspond to QTL for GLS resistance. CONCLUSIONS: QTL mapping of a RIL population from the subtropical maize parents CML444 and SC Malawi identified seven QTL for resistance to gray leaf spot disease caused by C.?zeina. These QTL together with QTL from eleven studies were allocated to bins on the IBM2005 map to provide a basis for comparison. Hotspots of GLS QTL were identified on chromosomes one, two, four, five and seven, with QTL in the current study overlapping with two of these. Two QTL from this study did not overlap with previously reported QTL.

Project description:BackgroundGray leaf spot (GLS), which is caused by the necrotrophic fungi Cercospora zeae-maydis and Cercospora zeina, is one of the most impactful diseases in maize worldwide. The aim of the present study is to identify the resistance genes and understand the molecular mechanisms for GLS resistance.ResultsTwo cultivars, 'Yayu889' and 'Zhenghong532,' which are distinguished as resistant and susceptible cultivars, respectively, were challenged with the GLS disease and a RNA-seq experiment was conducted on infected plants at 81, 89, 91, and 93 days post planting (dap). Compared with the beginning stage at 81 dap, 4666, 1733, and 1166 differentially expressed genes (DEGs) were identified at 89, 91, and 93 dap, respectively, in 'Yayu889,' while relatively fewer, i.e., 4713, 881, and 722 DEGs, were identified in 'Zhenghong532.' Multiple pathways involved in the response of maize to GLS, including 'response to salicylic acid,' 'protein phosphorylation,' 'oxidation-reduction process,' and 'carotenoid biosynthetic process,' were enriched by combining differential expression analysis and Weighted Gene Co-expression Network Analysis (WGCNA). The expression of 12 candidate resistance proteins in these pathways were quantified by the multiple reaction monitoring (MRM) method. This approach identified two candidate resistance proteins, a calmodulin-like protein and a leucine-rich repeat receptor-like protein kinase with SNPs that were located in QTL regions for GLS resistance. Metabolic analysis showed that, compared with 'Zhenghong532,' the amount of salicylic acid (SA) and total carotenoids in 'Yayu889' increased, while peroxidase activity decreased during the early infection stages, suggesting that increased levels of SA, carotenoids, and reactive oxygen species (ROS) may enhance the defense response of 'Yayu889' to GLS.ConclusionBy combining transcriptome and proteome analyses with comparisons of resistance QTL regions, calmodulin-like protein and leucine-rich repeat receptor-like protein kinase were identified as candidate GLS resistance proteins. Moreover, we found that the metabolic pathways for ROS, SA, and carotenoids are especially active in the resistant cultivar. These findings could lead to a better understanding of the GLS resistance mechanisms and facilitate the breeding of GLS-resistant maize cultivars.

Dataset Information

Deep Learning Diagnostics of Gray Leaf Spot in Maize under Mixed Disease Field Conditions.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets