Utilizing comparative genomic resources coupled with rapid, low cost targeted resequencing to develop robust diagnostic tools for food security

Each year, thousands of exotic and invasive species, particularly insects, plants, or diseases, are intercepted at ports of entry or detected in the environment. Crucial to management and eradication of these species, and... [ view full abstract ]

Each year, thousands of exotic and invasive species, particularly insects, plants, or diseases, are intercepted at ports of entry or detected in the environment. Crucial to management and eradication of these species, and prevention of their establishment, is rapid species identification, and determination of source of invasive material.  Our goal was to develop a rapid, straightforward tool for determining the species and source population of invasive Tephritid fruit flies that are commonly detected in California, Florida, and South Texas.  For many species, clear morphological tools for discriminating species do not exist, particularly at the immature level.  Here we present an approach for phylogenomic locus selection in these species that takes advantage of a variety of genomic and transcriptomic data sources, focusing on conserved exonic regions in orthologous genes.  This approach yielded a phylogenenomic tool far exceeding the resolution of existing approaches, such as target enrichment approaches or traditional marker sequencing, as loci are uniquely selected based off of their diagnostic utility.  Utilizing reduced representation sequencing of individual flies from through the geographic range, diagnostic loci for discriminating populations of specific invasive species were also developed.  To allow for rapid and low cost re-sequencing of these species and population level markers, we employed a highly multiplex, single tube amplicon sequencing approach, allowing targeting of thousands of loci at once, coupled with high throughput sequencing.  The wetlab and analysis pipeline developed can analyze hundreds of individuals at a time, and return taxonomic and population level assignment in a matter of three days from sample collection.  Our approach provides a novel way to combine diverse genomic and transcriptomic data sources, particularly when at least one well-annotated data source is available, and can rapidly develop diagnostic tools for non-model systems that are scalable, cost-effective, and robust.