Denoising Diffusion Probabilistic Models and Transfer Learning for citrus disease diagnosis

Problems

Plant Disease diagnosis based on deep learning mechanisms has been extensively studied and applied. However, the complex and dynamic agricultural growth environment results in significant variations in the distribution of state samples, and the lack of sufficient real disease databases weakens the information carried by the samples, posing challenges for accurately training models.

Aim

This paper aims to test the feasibility and effectiveness of Denoising Diffusion Probabilistic Models (DDPM), Swin Transformer model, and Transfer Learning in diagnosing citrus diseases with a small sample.

Methods

Two training methods are proposed: The Method 1 employs the DDPM to generate synthetic images for data augmentation. The Swin Transformer model is then used for pre-training on the synthetic dataset produced by DDPM, followed by fine-tuning on the original citrus leaf images for disease classification through transfer learning. The Method 2 utilizes the pre-trained Swin Transformer model on the ImageNet dataset and fine-tunes it on the augmented dataset composed of the original and DDPM synthetic images.

Results and conclusion

The test results indicate that Method 1 achieved a validation accuracy of 96.3%, while Method 2 achieved a validation accuracy of 99.8%. Both methods effectively addressed the issue of model overfitting when dealing with a small dataset. Additionally, when compared with VGG16, EfficientNet, ShuffleNet, MobileNetV2, and DenseNet121 in citrus disease classification, the experimental results demonstrate the superiority of the proposed methods over existing approaches to a certain extent.

Development of a Duplex qPCR System for Detection and Quantification of the Two Canola Blackleg Pathogens Leptosphaeria biglobosa and L. maculans

Two probe-based qPCR systems, namely P-Lb and P-Lm, specific to the canola blackleg pathogens Leptosphaeria biglobosa and L. maculans, respectively, were developed, and their efficiencies were tested. Each of the two systems targets a single-copy gene exclusively present in the corresponding species. The specificities of the two systems on the species level and their ubiquities on the subspecies level were confirmed by in silico sequence analyses and testing on L. biglobosa (17 strains), L. maculans (10 strains), and other plant pathogens (31 species). For sensitivities, the two systems were tested on synthesized DNA fragments (gBlock) of the targeted regions, from which a standard curve was generated for each system. In addition, standard curves were also generated on gBlocks for duplex qPCR in which the two systems were used in the same reaction. The two systems were further tested in both singleplex and duplex qPCR on DNA samples extracted from fungal spores, inoculated canola cotyledons, and naturally infected canola stubble samples collected from commercial fields. Our data indicated that the two systems are specific to L. biglobosa and L. maculans, respectively, and one reaction could detect as few as 200 spores of either species. When used in duplex qPCR on DNA samples with various origins, the two systems generated similar results as in singleplex qPCR. The duplex qPCR system, along with the sample preparation and DNA extraction specified in this study, constituted a first-reported duplex qPCR protocol for detection and quantification of the two blackleg pathogens from field samples.

Enhancing host-pathogen phenotyping dynamics: early detection of tomato bacterial diseases using hyperspectral point measurement and predictive modeling

Early diagnosis of plant diseases is needed to promote sustainable plant protection strategies. Applied predictive modeling over hyperspectral spectroscopy (HS) data can be an effective, fast, cost-effective approach for improving plant disease diagnosis. This study aimed to investigate the potential of HS point-of-measurement (POM) data for in-situ, non-destructive diagnosis of tomato bacterial speck caused by Pseudomonas syringae pv. tomato (Pst), and bacterial spot, caused by Xanthomonas euvesicatoria (Xeu), on leaves (cv. cherry). Bacterial artificial infection was performed on tomato plants at the same phenological stage. A sensing system composed by a hyperspectral spectrometer, a transmission optical fiber bundle with a slitted probe and a white light source were used for spectral data acquisition, allowing the assessment of 3478 spectral points. An applied predictive classification model was developed, consisting of a normalizing pre-processing strategy allied with a Linear Discriminant Analysis (LDA) for reducing data dimensionality and a supervised machine learning algorithm (Support Vector Machine - SVM) for the classification task. The predicted model achieved classification accuracies of 100% and 74% for Pst and Xeu test set assessments, respectively, before symptom appearance. Model predictions were coherent with host-pathogen interactions mentioned in the literature (e.g., changes in photosynthetic pigment levels, production of bacterial-specific molecules, and activation of plants' defense mechanisms). Furthermore, these results were coherent with visual phenotyping inspection and PCR results. The reported outcomes support the application of spectral point measurements acquired in-vivo for plant disease diagnosis, aiming for more precise and eco-friendly phytosanitary approaches.

Advancing the automation of plant nucleic acid extraction for rapid diagnosis of plant diseases in space

Human space exploration missions will continue the development of sustainable plant cultivation in what are obviously novel habitat settings. Effective pathology mitigation strategies are needed to cope with plant disease outbreaks in any space-based plant growth system. However, few technologies currently exist for space-based diagnosis of plant pathogens. Therefore, we developed a method of extracting plant nucleic acid that will facilitate the rapid diagnosis of plant diseases for future spaceflight applications. The microHomogenizer^™ from Claremont BioSolutions, originally designed for bacterial and animal tissue samples, was evaluated for plant-microbial nucleic acid extractions. The microHomogenizer^™ is an appealing device in that it provides automation and containment capabilities that would be required in spaceflight applications. Three different plant pathosystems were used to assess the versatility of the extraction process. Tomato, lettuce, and pepper plants were respectively inoculated with a fungal plant pathogen, an oomycete pathogen, and a plant viral pathogen. The microHomogenizer^™, along with the developed protocols, proved to be an effective mechanism for producing DNA from all three pathosystems, in that PCR and sequencing of the resulting samples demonstrated clear DNA-based diagnoses. Thus, this investigation advances the efforts to automate nucleic acid extraction for future plant disease diagnosis in space.

New-Generation Sequencing Technology in Diagnosis of Fungal Plant Pathogens: A Dream Comes True?

The fast and continued progress of high-throughput sequencing (HTS) and the drastic reduction of its costs have boosted new and unpredictable developments in the field of plant pathology. The cost of whole-genome sequencing, which, until few years ago, was prohibitive for many projects, is now so affordable that a new branch, phylogenomics, is being developed. Fungal taxonomy is being deeply influenced by genome comparison, too. It is now easier to discover new genes as potential targets for an accurate diagnosis of new or emerging pathogens, notably those of quarantine concern. Similarly, with the development of metabarcoding and metagenomics techniques, it is now possible to unravel complex diseases or answer crucial questions, such as "What's in my soil?", to a good approximation, including fungi, bacteria, nematodes, etc. The new technologies allow to redraw the approach for disease control strategies considering the pathogens within their environment and deciphering the complex interactions between microorganisms and the cultivated crops. This kind of analysis usually generates big data that need sophisticated bioinformatic tools (machine learning, artificial intelligence) for their management. Herein, examples of the use of new technologies for research in fungal diversity and diagnosis of some fungal pathogens are reported.

Nanomaterials for Agricultural and Ecological Defense Applications: Active Agents and Sensors

The world we live in today is overpopulated with an unprecedented number of people competing for fewer and fewer precious resources. The struggle to efficiently steward and manage these resources is a global problem in need of concrete and urgent solutions. Nanomaterials have driven innovation in diverse industrial sectors including military, aviation, electronic, and medical among others. Nanoscale materials possess unique surfaces and exquisite opto-electronic properties that make them uniquely suited to environmental, biological, and ecological defense applications. A tremendous upsurge of research activity in these areas is evident from the exponential increase in publications worldwide. Here we review recent applications of nanomaterials toward soil health and management, abiotic and biotic stress management, plant defense, delivery of the RNA Interference (RNAi), plant growth, manufacture of agro-products, and ecological investigations related to farming. For example, nanomaterial constructs have been used to counter environmental stresses and in plant defense and disease diagnosis. Nanosensor chemistries have been developed to monitor water quality and measure specific pollutant levels. Specific nanomaterials such as silver, iron oxide, and zinc oxide proffer protection to plants from pathogens. This review describes progress in nanomaterial-based agricultural and ecological defense and seeks to identify factors that would enable their wider commercialization and deployment. This article is categorized under: Diagnostic Tools > Biosensing Toxicology and Regulatory Issues in Nanomedicine > Toxicology of Nanomaterials Diagnostic Tools > Diagnostic Nanodevices.

Rapid In Situ Analysis of Plant Emission for Disease Diagnosis Using a Portable Gas Chromatography Device

We developed and applied a fully automated portable gas chromatography (GC) device for rapid and in situ analysis of plant volatile organic compounds (VOCs) to examine plant health status. A total of 42 emission samples were collected over a period of 5 days from 10 milkweed ( Asclepias syriaca) plants, half of which were infested by aphids. Thirty-five VOC peaks were separated and detected in 8 min. An algorithm based on machine learning, principal component analysis, and linear discriminant analysis was developed to evaluate the GC results. We found that our device and algorithm are able to distinguish between the undamaged control and the aphid-infested milkweeds with an overall accuracy of 90-100% within 48-72 h of the attack. Such rapid in situ detection of insect attack attests to the great potential of VOC monitoring in plant health management.

Differential Usefulness of Nine Commonly Used Genetic Markers for Identifying Phytophthora Species

The genus Phytophthora is agriculturally and ecologically important. As the number of Phytophthora species continues to grow, identifying isolates in this genus has become increasingly challenging even by DNA sequencing. This study evaluated nine commonly used genetic markers against 154 formally described and 17 provisionally named Phytophthora species. These genetic markers were the cytochrome-c oxidase 1 (cox1), internal transcribed spacer region (ITS), 60S ribosomal protein L10, beta-tubulin (β-tub), elongation factor 1 alpha, enolase, heat shock protein 90, 28S ribosomal DNA, and tigA gene fusion protein (tigA). As indicated by species distance, cox1 had the highest genus-wide resolution, followed by ITS, tigA, and β-tub. Resolution of these four markers also varied with (sub)clade. β-tub alone could readily identify all species in clade 1, cox1 for clade 2, and tigA for clades 7 and 8. Two or more genetic markers were required to identify species in other clades. For PCR consistency, ITS (99% PCR success rate) and β-tub (96%) were easier to amplify than cox1 (75%) and tigA (71%). Accordingly, it is recommended to take a two-step approach: classifying unknown Phytophthora isolates to clade by ITS sequences, as this marker is easy to amplify and its signature sequences are readily available, then identifying to species by one or more of the most informative markers for the respective (sub)clade.