Viroscope: Honing virus detection for the global plant trade

• The destruction of plant crops from the propagation of plant pathogens, such as viruses, is a serious threat to global food security and human perpetuity.
• The world’s governing bodies lack the tools for standardised and accurate virus detection.
• The CEO and Founder of Multiplex startup, Dr Bernardo Pollak, and fellow researchers based in Chile have developed and tested a unique diagnostic pipeline, Viroscope, and demonstrated its functionality through a web application, Viroscope.io.
• The algorithm and cloud-based platform are intended to improve virus detection for governmental phytosanitary programmes and expedite the quarantine process for plant materials.

A staggering volume of plant crops are lost to plant pests and vigorous diseases caused by plant pathogens, referring to organisms such as bacteria, fungi, and viruses – a primary threat to global food security. According to estimates from the Food and Agriculture Organization of the United Nations, plant diseases account for the loss of up to 40% of global crop production annually, costing the agricultural sector more than US$ 220 billion. With the increase in average global temperatures spurred by climate change, a growing human population, an increase in consumer food demand, and the subsequent import and export of plant materials from the global plant trade, the threat of propagating plant pathogens is of palpable concern.

The pervading threat to global food security posed by plant viral pathogens obviates the need for effective and expedient governmental phytosanitary programmes. The main goal of phytosanitary programmes is to stop the spread of diseases and pests that may decimate crops, forests, and other natural ecosystems and promote secure global plant trade.

Improving virus detection

Dr Bernardo Pollak, along with fellow collaborative researchers in Chile, has devised a sophisticated algorithm, called Viroscope, that can help improve the accuracy of virus detection.

The unique diagnostic pipeline is supported by a cloud-based platform that broadens access to phytosanitary agencies across the globe through the provision of high-throughput sequencing (HTS) data interpretation and consequent viral pathogen identification.

Plant diseases account for the loss of up to 40% of global crop production annually, costing the agricultural sector more than US$ 220 billion.

Although genetic analysis and detection tools such as HTS are hailed as a gold standard for diagnosis and are becoming more affordable, several hurdles lie in the way of the technique’s wider adoption. These include the necessity for bioinformatics knowledge and processing power, inconsistent bioinformatics pipeline metrics, and the need for consistent criteria for viral identification, particularly in circumstances of low sequencing coverage.

Pipeline overview and validation

In their 2022 paper, Pollak and his research team describe the Viroscope algorithm. It is comprised of four key steps: detection, analysis, interpretation, and diagnosis. First off, we begin with the biological material, which in this case are field samples from four sweet cherry plants. Ribose nucleic acid or RNA – a genetic component found in many organisms including viruses – is then extracted for further processing and analysis. Extracted RNA undergoes HTS, a contemporary technology for rapidly sequencing DNA or RNA molecules. This technique enables the parallel sequencing of millions to billions of RNA or DNA fragments.

In the analysis phase, Viroscope employs the process of read assignment to assign or map specific genetic sequences or fragments (also called reads, produced from HTS sequencing) to a reference genome or, in this case, a curated database of viral targets. The team used three read assignment softwares to identify genomic or transcriptome regions from which each read was extracted. Only high-quality reads are retained for further processing in the Viroscope pipeline. Mapped RNA reads are gathered and used for creating longer contiguous sequences to calculate viral genome assembly coverage (VGAC, representing the level of recovery of the viral genomes following sequencing).

Following this stage, the pipeline searches for the presence of protein sequences associated with virus replication (namely those proteins encoding polymerase and replicase enzymes) using an HTS alignment tool called Diamond with the help of Viral DataBase (RVDB)-prot, a database that contains a compilation of viral proteins to aid identification. Replicases are enzymes necessary for the replication of RNA viruses. They support the reproduction of the virus inside host cells by producing new viral RNA genomes.

Finally, these measurements are employed to interpret viral presence in accordance with predetermined diagnostic cutoffs. Sweet cherry field samples, a simulation dataset, a mutation dataset, and external published datasets were utilised to validate the Viroscope pipeline.

HTS data and VGAC-backed virus detection

Pollak and his team found VGAC to be superior over read assignment for virus detection. False positives can result from read assignment, especially when host sequences include viral origins, such as endogenous viral components. For instance, the integration of reverse transcriptase domains by badnaviruses into the host genome isn’t always a sign of an ongoing infection. Badnaviruses are plant viruses that may induce disease in a variety of crops, impacting growth and yield. Locating badnavirus replicases is important since they may well be active and provide a systemic viral infection risk.

Furthermore, the researchers show that HTS-based diagnostics exhibit greater sensitivity than real-time polymerase chain reaction (RT-PCR) – another sequencing method – and can identify a wider variety of viruses.

Overall, the researchers demonstrate the significance of employing replicase identification and VGAC for enhancing the accuracy of virus detection through the Viroscope pipeline. Compared to traditional techniques like RT-PCR, Pollak and fellow researchers believe that their methodology offers greater sensitivity and accuracy concerning HTS data-backed virus diagnosis.

A large-scale field study

Alongside Pollak, Morgante and colleagues (Morgante, et al, 2023) conducted a large-scale study (with 220 samples) employing the Viroscope.io web service for swift virus detection using HTS data to support efficient phytosanitary quarantine operations. Leaf samples from the SAG (Servicio Agrícola Ganadero) phytosanitary agency came from a variety of plants, including apple, plum, sweet cherry olive, citrus, and avocado. The study contrasts Viroscope’s functional annotation with traditional methods utilised in phytosanitary quarantines, emphasising the additional insights it offers, particularly in situations where viral abundance is low.

Pollak and fellow researchers believe that their methodology offers greater sensitivity and accuracy concerning HTS data-backed virus diagnosis.

Viroscope.io was used in the study to examine 145 plants for viruses; however, no quarantine viruses or viroids were discovered. The platform allows users to upload their data, initiate analyses, and select a predetermined panel of target viruses for detection.

HTS proved effective, with 2.75% controlled and 29.6% emerging viruses. Viroscope exceeded conventional technologies, providing exact detection, cost, and time reductions, as well as insight into viral physiology. Viroscope has been validated for real-life situations and is set to become a valuable tool for governmental phytosanitary programmes, one that can help expedite the quarantine process and safeguard global food security.