This project will improve the way pepper varieties are selected for flavor, aroma, and consumer preference

The Department of Agriculture National Institute of Food and Agriculture awarded West Virginia State University Research and Development Corporation $299,240 on September 1, 2026, under the Resident Instruction, Agriculture, and Food Science Facilities and Equipment Grants program (CFDA 10.308) to improve pepper breeding through flavor and aroma selection.

The project will study 244 Capsicum chinense accessions—including habanero and Scotch bonnet varieties—to understand how genetic and chemical differences influence flavor, aroma, and consumer preference. The work combines genomic analysis with measurements of volatile and nonvolatile compounds that contribute to taste and smell. Consumer panels will evaluate representative peppers for aroma, flavor, heat, and overall acceptability, describing perceived sensory characteristics. Machine learning models will then predict consumer preference from genomic and metabolomic data, enabling breeders to identify promising pepper lines without relying solely on repeated sensory evaluations. The project will deliver a user-friendly prediction tool providing expected flavor and aroma characteristics from genetic and metabolite data, and will evaluate useful markers and flavor-related traits identified in C. chinense across other cultivated pepper species including C. annuum, C. frutescens, and C. baccatum to identify germplasm for future breeding.

Period of performance runs from September 1, 2026, through August 31, 2028. Work is performed in Dunbar, West Virginia. West Virginia State University Research and Development Corporation is a technology transfer and research arm of West Virginia State University, a 1890 Land Grant College and Historically Black College or University.

Capsicum chinense, which includes habanero, scotch bonnet, and other specialty peppers, is valued not only for heat but also for its distinctive fruity, floral, citrus, and smoky flavors. However, most pepper breeding programs have traditionally focused more on yield, disease resistance, and pungency than on the complex traits that determine flavor quality. This project will study a diverse collection of 244 c. Chinense accessions to understand how differences in their DNA and chemical composition influence flavor and aroma. The project will combine genomic information with measurements of volatile and nonvolatile compounds that contribute to taste and smell. Consumers will also evaluate representative peppers for aroma, flavor, heat, and overall acceptability and will describe the sensory characteristics they perceive. These data will be combined to identify the genes, genetic markers, and metabolites most strongly associated with desirable flavor profiles. Machine learning methods will then be used to develop models that can predict consumer preference from genomic and metabolomic information. The goal is to give breeders a practical way to identify promising pepper lines without relying only on repeated sensory evaluations. The project will also develop a user friendly prediction tool that can provide expected flavor and aroma characteristics from genetic and metabolite data. In addition, the most useful markers and flavor related traits identified in c. Chinense will be evaluated in other cultivated pepper species, including c. Annuum, c. Frutescens, and c. Baccatum, to identify promising material for future breeding. The project is expected to provide breeders with improved markers, useful germplasm, predictive tools, and a better understanding of the biological factors that shape consumer preference. These resources will support the development of pepper cultivars with distinctive flavor, desirable aroma, moderate pungency, and improved market appeal. The project will also provide research and training opportunities for students and early career scientists in plant breeding, genomics, metabolomics, sensory science, and machine learning, helping build capacity for future data driven crop improvement.