Agricultural high throughput diagnostics need to be fast, accurate and have multiplexing capacity. Metagenomic sequencing is being widely evaluated for plant and animal diagnostics. Bioinformatic analysis of metagenomic sequence data has been a bottleneck for diagnostic analysis due to the size of the data files. Most available tools for analyzing high-throughput sequencing (HTS) data require that the user have computer coding skills and access to high-performance computing. To overcome constraints to most sequencing-based diagnostic pipelines today, we have developed Microbe Finder (MiFi^®). MiFi^® is a web application for quick detection and identification of known pathogen species/strains in raw, unassembled HTS metagenomic data. HTS-based diagnostic tools developed through MiFi^® must pass rigorous validation, which is outlined in this manuscript. MiFi^® allows researchers to collaborate in the development and validation of HTS-based diagnostic assays using MiProbe™, a platform used for developing pathogen-specific e-probes. Validated e-probes are made available to diagnosticians through MiDetect™. Here we describe the e-probe development, curation and validation process of MiFi^® using grapevine pathogens as a model system. MiFi^® can be used with any pathosystem and HTS platform after e-probes have been validated.

Early stage infections caused by fungal/oomycete spores may not be detected until signs or symptoms develop. Serological and molecular techniques are currently used for detecting these pathogens. Next-generation sequencing (NGS) has potential as a diagnostic tool, due to the capacity to target multiple unique signature loci of pathogens in an infected plant metagenome. NGS has significant potential for diagnosis of important eukaryotic plant pathogens. However, the assembly and analysis of huge amounts of sequence is laborious, time consuming, and not necessary for diagnostic purposes. Previous work demonstrated that a bioinformatic tool termed Electronic probe Diagnostic Nucleic acid Analysis (EDNA) had potential for greatly simplifying detecting fungal and oomycete plant pathogens in simulated metagenomes. The initial study demonstrated limitations for detection accuracy related to the analysis of matches between queries and metagenome reads. This study is a modification of EDNA demonstrating a better accuracy for detecting fungal and oomycete plant pathogens.

High-throughput sequencing (HTS) is becoming the new norm of diagnostics in plant quarantine settings. HTS can be used to detect, in theory, all pathogens present in any given sample. The technique’s success depends on various factors, including methods for sample management/preparation and suitable bioinformatic analysis. The Limit of Detection (LoD) of HTS for plant diagnostic tests can be higher than that of PCR, increasing the risk of false negatives in the case of low titer of the target pathogen. Several solutions have been suggested, particularly for RNA viruses, including rRNA depletion of the host, dsRNA, and siRNA extractions, which increase the relative pathogen titer in a metagenomic sample. However, these solutions are costly and time-consuming. Here we present a faster and cost-effective alternative method with lower HTS-LoD similar to or lower than PCR. The technique is called TArget-SPecific Reverse Transcript (TASPERT) pool. It relies on pathogen-specific reverse primers, targeting all RNA viruses of interest, pooled and used in double-stranded cDNA synthesis. These reverse primers enrich the sample for only pathogens of interest. Evidence on how TASPERT is significantly superior to oligodT, random 6-mer, and 20-mer in generating metagenomic libraries containing the pathogen of interest is presented in this proof of concept.

The advancement in high-throughput sequencing (HTS) technology allows the detection of pathogens without the need for isolation or template amplification. Plant regulatory agencies worldwide are adopting HTS as a prescreening tool for plant pathogens in imported plant germplasm. The technique is a multipronged process and, often, the bioinformatic analysis complicates detection. Previously, we developed E-probe diagnostic nucleic acid analysis (EDNA), a bioinformatic tool that detects pathogens in HTS data. EDNA uses custom databases of signature nucleic acid sequences (e-probes) to reduce computational effort and subjectivity when determining pathogen presence in a sample. E-probes of Pythium ultimum and Phytophthora ramorum were previously validated only using simulated HTS data. However, HTS samples generated from infected hosts or pure culture may vary in pathogen concentration, sequencing bias, and data quality, suggesting that each pathosystem requires further validation. Here, we used metagenomic and genomic HTS data generated from infected hosts and pure culture, respectively, to further validate and curate e-probes of Pythium ultimum and Phytophthora ramorum. E-probe length was found to be a determinant of diagnostic sensitivity and specificity; 80-nucleotide e-probes increased the diagnostic specificity to 100%. Curating e-probes to increase specificity affected diagnostic sensitivity only for 80-nucleotide Pythium ultimum e-probes. Comparing e-probes with alternative databases and bioinformatic tools in their speed and ability to find Pythium ultimum and Phytophthora ramorum demonstrated that, although pathogen sequence reads were detected by other methods, they were less specific and slower when compared with e-probes.

E-probe Diagnostic for Nucleic acid Analysis (EDNA) is a bioinformatic tool originally developed to detect plant pathogens in metagenomic databases. However, enhancements made to EDNA increased its capacity to conduct hypothesis directed detection of specific gene targets present in transcriptomic databases. To target specific pathogenicity factors used by the pathogen to infect its host or other targets of interest, e-probes need to be developed for transcripts related to that function. In this study, EDNA transcriptomics (EDNAtran) was developed to detect the expression of genes related to aflatoxin production at the transcriptomic level. E-probes were designed from genes up-regulated during A. flavus aflatoxin production. EDNAtran detected gene transcripts related to aflatoxin production in a transcriptomic database from corn, where aflatoxin was produced. The results were significantly different from e-probes being used in the transcriptomic database where aflatoxin was not produced (atoxigenic AF36 strain and toxigenic AF70 in Potato Dextrose Broth).