Innovations in genomics and phenomics in breeding program of tilapia

• Grant number: 24/11172-1
• Support Opportunities: Regular Research Grants
• Start date: July 01, 2025
• End date: June 30, 2028
• Field of knowledge: Agronomical Sciences - Fishery Resources and Fishery Engineering - Aquaculture
• Principal Investigator: Diogo Teruo Hashimoto
• Grantee: Diogo Teruo Hashimoto
• Host Institution: Centro de Aquicultura (CAUNESP). Universidade Estadual Paulista (UNESP). Campus de Jaboticabal. Jaboticabal , SP, Brazil
• Associated researchers: Guilherme Jordão de Magalhães Rosa

Abstract

Currently, new technologies are necessary to support the growing volume of fish production and the demand for fish consumption, as a means of ensuring food security and bioeconomy. Genetic improvement programs are among the primary methods to achieve efficient aquaculture industry, especially with investments in genomic and phenomic innovations. The study hypothesis aims to test whether these technological innovations can be applied to select tilapia genotypes with improved fillet yield, enabling accurate predictions of estimated breeding value (EBV). Genomic selection studies, through genotype imputation, have shown to increase the accuracy EBV predictions compared to pedigree-based methods (traditional selection); therefore, this strategy will be tested in the current project to accelerate genetic progress in fillet yield, which is considered one of the phenotypes that are difficult or impossible to directly measure in selection candidates, thus implying the sacrifice of the animal and the discard of a potential breeder. Alongside the genomic era, high-throughput phenotyping technologies (phenomics) have recently become available, especially those utilizing artificial intelligence and machine learning (deep learning), which will create opportunities to incorporate big data and precision aquaculture into genetic improvement programs. The analysis of 3D images and computer vision systems (CVS), through deep learning, will be tested as a method to objectively and efficiently measure fillet yield, thereby facilitating the inclusion of this phenotype as a selection criterion. The expected outcomes are: 1) to generate a low-cost genomic tool for selecting superior tilapia genotypes; 2) a CVS (software) for obtaining real-time and non-invasive fillet yield measurements; and 3) to demonstrate the potential to improve tilapia fillet yield through the integration of technological innovations. (AU)