Project Number: 2094-43000-008-05-T
Project Type: Trust Fund Cooperative Agreement
Start Date: Apr 1, 2019
End Date: Mar 31, 2022
1) Sequence genomes using Honaas’ workflow to efficiently discover variety specific gene forms for ‘Honeycrisp’ (ready in queue), ‘Cosmic Crisp’ (ready in queue), and ‘Gala’ (ready in year 1), 2) refine biomarker discovery pipeline using machine learning algorithms 3) begin validation of biomarkers by designing gene tests for use in multi-lot surveys.
Global-scale gene activity measurements (i.e. measure activity of all ~40,000 genes simultaneously, or RNA-Seq) require prior accurate knowledge of the genes. Typically, this is a reference genome – in the case of apple the only reference is Golden Delicious. Because previous work, including published and ongoing work, has shown that differences between genes across cultivars reduces the quality of RNA-Seq, we will begin with sequencing of the genomes of 3 apple varieties. Cooperating with Penn State, ARS will use an efficient hybrid genome sequencing approach (2nd and 3rd gen. technologies) to assemble genomes that represent the high complexity, gene-rich regions of 3 apple cultivars: Gala, Honey Crisp, and Cosmic Crisp. This approach is an efficient way to discover the full set of cultivar-specific genes that we can use as a reference for RNA-Seq. Having a genome that perfectly matches the gene activity data improves the quality of the measurements and thus the quality of downstream analyses. Plus, this generates a resource for future work in these economically important cultivars, including comparative genomics that can be leveraged to develop postharvest tools useful for across varieties in an increasingly variety-diverse marketplace. Concurrent to genome sequencing and assembly in year 1, and then subsequently in years 2 and 3, ARS will setup and run the postharvest experiments each year (for discovery and then validation), cryogenically preserve the fruit tissue, and generate the RNA-Seq data for downstream analyses with the new genomes. ARS will provide fruit quality data for all corresponding gene activity measurements. The downstream analysis includes a biomarker discovery step (gene activity correlation, machine learning algorithms, comparative network analysis) that will produce a list of potential biomarker genes associated with fruit quality traits of interest. ARS will validate the biomarker candidate genes in a multi-lot, multi-year scheme to evaluate their usefulness to predict future fruit quality.