GPSS Research Symposium
Permanent URI for this collection
Programs and presentations from the Graduate and Professional Student Senate (formerly GSA) Research Symposium, held every spring.
Browse
Browsing GPSS Research Symposium by Author "Mechan Llontop, Marco E."
Now showing 1 - 1 of 1
Results Per Page
Sort Options
- Strain-level identification of tomato pathogens from metagenomic sequences obtained with the ONT MinIONSharma, Parul; Mechan Llontop, Marco E.; Aguilera Flores, Marcela; Li, Song; Vinatzer, Boris A. (Virginia Tech, 2020-03-25)Early detection and correct diagnosis of plant diseases is an essential component of sustainable production of food and other plant-derived products. Although molecular technologies are available, many of them are either slow because they depend on culturing the pathogen first, are limited to specific pathogen species and thus cannot detect any newly emerging diseases, or have low resolution. With recent advances in sequencing technologies, it has become possible to sequence the DNA of an entire plant sample, called the metagenome, at relatively low cost and with relatively easy and fast protocols using the Oxford Nanopore Technologies (ONT) MinIONTM device. MinIONTM software What’s in my pot (WIMP) offers read-based taxonomic identification from the metagenome. In this study, we have used the MinIONTM device to sequence laboratory-inoculated tomato plants and field samples of infected tomato plants to establish the efficiency of WIMP in identifying the underlying plant pathogens. The taxonomic classifications, at the species-level, from WIMP were compared with the results from the third party Sourmash and MetaMaps tools. Since species-level identification is not always sufficient, for example, when tracking pathogen dissemination pathways, custom reference libraries were used to attempt strain-level classification with Sourmash and MetaMaps as well as identification with the LINbase Web service based on metagenome-assembled genomes (MAGs). Our study showed that reliable species-level identification is possible with either WIMP, Sourmash, or MetaMaps. There is the potential for strain-level accuracy, however improvements in the error rate of the MinIONTM and availability of appropriate reference databases is necessary.