Interpretive Summary: Artificial intelligence in animal breeding and genetics: applications, opportunities, and challenges

By: Gota Morota, Hyo-Jun Lee, Seung-Hwan Lee, Hee-Bok Park, Akio Onogi

Implications

• Applications of artificial intelligence in animal breeding and genetics can be broadly categorized into two areas: phenotype generation and predictive genetic modeling.
• Previous studies have primarily focused on using artificial intelligence for digital phenotyping, whereas its integration with genomic prediction remains in its early stages.
• Artificial intelligence, in the form of foundation models, is becoming an indispensable tool for generating animal phenotypes from image and sensor data, with highly promising future prospects.
• The application of artificial intelligence to tabular data, such as single nucleotide polymorphism datasets, continues to pose significant challenges in the context of genomic prediction.

Introduction

Animal science has become a data-rich field, particularly within the area of animal breeding and genetics, where millions of phenotyped and genotyped animals are now available across livestock species databases. Genomic selection is now routinely used for genetic evaluations in many livestock species. Meanwhile, artificial intelligence (AI) has become a major focus in recent years, appearing in nearly every aspect of our daily lives. Animal breeding and genetics is no exception, and questions remain regarding where and how AI can be effectively integrated into this field. Broadly speaking, the role of AI in animal breeding and genetics is twofold: phenotype generation and predictive genetic modeling. Although the utility of AI in animal breeding and genetics has been reviewed previously, most existing studies primarily highlight its applications in digital phenotyping (Morota et al., 2018; Koltes et al., 2019; Grohmann and Decker, 2024; Klingström et al., 2025). This is because, by harnessing digital data sources, such as images, video recordings, and sensor signals, digital phenotyping enables efficient prediction and classification of animal phenotypes. However, the convergence of AI and genomic prediction models is still in its early stages. In light of this, this feature article emphasizes not only the application of AI in animal phenotyping but also the exploration of phenotype–genotype relationships through the lens of AI.

Read more in Animal Frontiers: From Isolated Data to Integrated Ecosystems.