Aquaculture 2025

March 6 - 10, 2025

New Orleans, Louisiana USA

MACHINE LEARNING ANALYSIS OF GENE EXPRESSION PATTERNS IN STRIPED BASS, WHITE BASS, AND THEIR HYBRID FOLLOWING BACTERIAL INFECTIONS

Linnea K. Andersen*, Jason W. Abernathy, Bradley D. Farmer, Miles D. Lange, Nithin M. Sankappa, Matthew E. McEntire, and Steven D. Rawles

*Aquatic Animal Health Research Unit (AAHRU), Agricultural Research Service (ARS), United States Department of Agriculture (USDA), Auburn, AL, USA. Linnea.Andersen@usda.gov

 



Striped bass (SB, Morone saxatilis) and white bass (WB, Morone chrysops) are important aquaculture species and parental contributors to hybrid striped bass (HSB), the fourth largest finfish aquaculture industry in the U.S. Recent studies have examined gene expression changes in the gill and spleen of these fish over time following infection with three pathogenic bacteria that significantly impact cultured and wild populations: Aeromonas veronii, Flavobacterium columnare, and Streptococcus iniae. These studies revealed differences in the expression of pattern recognition receptors (PRRs), cytokines, apoptotic factors, and genes involved in metabolism and bioenergetics, which, when combined with observed survival data, offer insights into the resistance or susceptibility of each group to infection.

In this work, an ensemble machine learning approach was applied to these time series gene expression data to detect key immune response patterns in both timing and magnitude across the causative bacterial agents and SB, WB, and HSB groups. The comparison of SB and WB provides valuable insight into the parental contributions to the hybrid vigor (heterosis) observed in HSB in response to some, but not all, pathogens. Furthermore, incorporating survival data into the machine learning models enables the development of predictive survival models that can be expanded with future data and used for selective breeding and vaccine or treatment development.