No abstract

We report here the biological and molecular attributes of a novel dsRNA mycovirus designated Rhizoctonia solani partitivirus 2 (RsPV2) from strain GD-11 of R. solani AG-1 IA, the causal agent of rice sheath blight. The RsPV2 genome comprises two dsRNAs, each possessing a single ORF. Phylogenetic analyses indicated that this novel virus species RsPV2 showed a high sequence identity with the members of genus Alphapartitivirus in the family Partitiviridae, and formed a distinct clade distantly related to the other genera of Partitiviridae. Introduction of purified RsPV2 virus particles into protoplasts of a virus-free virulent strain GD-118 of R. solani AG-1 IA resulted in a derivative isogenic strain GD-118T with reduced mycelial growth and hypovirulence to rice leaves. Taken together, it is concluded that RsPV2 is a novel dsRNA virus belonging to Alphapartitivirus, with potential role in biological control of R. solani.

The complete genome of a novel mycovirus, Rhizoctonia solani dsRNA virus 1 (RsRV1) was sequenced and analyzed. It is composed of two dsRNA genome segments, 2379 bp and 1811 bp in length, which were referred to as RsRV1-1 and RsRV1-2, respectively. RsRV1-1 contains a single open reading frame (ORF1), which has a conserved RNA-dependent RNA polymerase (RdRp) domain, whereas RsRV1-2 contains a single ORF2, which might encode a multifunctional protein. The genome organization of RsRV1 is similar to that of members of the family Partitiviridae. However, phylogenetic analysis indicated that RsRV1 formed a distinct clade together with three other unclassified viruses, suggesting that RsRV1 may belong to a new family of dsRNA mycoviruses. This is the first report of the full-length nucleotide sequence of a novel dsRNA mycovirus, RsRV1, infecting R. solani AG-1 IA strain B275, the causal agent of rice sheath blight.

The complete sequence and genome organization of a novel Endornavirus from the hypovirulent strain GD-2 of Rhizoctonia solani AG-1 IA, the causal agent of rice sheath blight, were identified using a deep sequencing approach and it was tentatively named as Rhizoctonia solani endornavirus 1 (RsEV1). It was composed of only one segment that was 19,936 bp in length and was found to be the longest endornavirus genome that has been reported so far. The RsEV1 genome contained two open reading frames (ORFs): ORF1 and ORF2. ORF1 contained a glycosyltransferase 1 domain and a conserved RNA-dependent RNA polymerase domain, whereas ORF2 encoded a conserved hypothetical protein. Phylogenetic analysis revealed that RsEV1 was phylogenetically a new endogenous RNA virus. A horizontal transmission experiment indicated that RsEV1 could be transmitted from the host fungal strain GD-2 to a virulent strain GD-118P and resulted in hypovirulence in the derivative isogenic strain GD-118P-V1. Metabolomic analysis showed that 32 metabolites were differentially expressed between GD-118P and its isogenic hypovirulent strain GD-118P-V1. The differential metabolites were mainly classified as organic acids, amino acids, carbohydrates, and the intermediate products of energy metabolism. Pathway annotation revealed that these 32 metabolites were mainly involved in pentose and glucuronate interconversions and glyoxylate, dicarboxylate, starch, and sucrose metabolism, and so on. Taken together, our results showed that RsEV1 is a novel Endornavirus, and the infection of virulent strain GD-118P by RsEV1 caused metabolic disorders and resulted in hypovirulence. The results of this study lay a foundation for the biocontrol of rice sheath blight caused by R. solani AG1-IA.

Recently, many Deep Learning models have been employed to classify different kinds of plant diseases, but very little work has been done for disease severity detection. However, it is more important to master the severities of plant diseases accurately and timely, as it helps to make effective decisions to protect the plants from being further infected and reduce financial loss. In this paper, based on the Huanglongbing (HLB)-infected leaf images obtained from PlantVillage and crowdAI, we created a dataset with 5,406 citrus leaf images infected by HLB. Then six popular models were trained to perform the severity detection of citrus HLB with the goal to find which types of models are more suitable to detect HLB severity with the same training circumstance. The experimental results show that the Inception_v3 model with epochs=60 can achieve higher accuracy than that of other models for severity detection with an accuracy of 74.38% due to its highly computational efficiency and small number of parameters. Additionally, aiming for evaluating whether the augmented data can contribute to improve the model learning performance, we adopted Generative Adversarial Networks (GANs) to augment the original training dataset up to two times itself. Finally, a new training dataset with 14,056 leaf images composed by the original training images and the augmented ones were used to train the Inception_v3. As a result, we achieved an accuracy of 92.60%, about 20% higher than that of model trained by the original training dataset, which suggested that the GANs-based data augmentation is very useful to improve the model learning performance.