Location: Livestock Nutrient Management Research
Title: Computational approaches for enteric methane mitigation research: From fermi calculations to artificial intelligence paradigmsAuthor
![]() |
CHOWDHURY, RATUL - Iowa State University |
![]() |
Frazier, Anthony |
![]() |
Koziel, Jacek |
![]() |
THOMPSON, LOGAN - Kansas State University |
![]() |
Beck, Matthew |
Submitted to: Animal Frontiers
Publication Type: Literature Review Publication Acceptance Date: 7/20/2024 Publication Date: 1/4/2025 Citation: Chowdhury, R., Frazier, A.N., Koziel, J.A., Thompson, L.R., Beck, M.R. 2024. Computational approaches for enteric methane mitigation research: From fermi calculations to artificial intelligence paradigms. Animal Frontiers. 14(6):33-41. https://doi.org/10.1093/af/vfae025. DOI: https://doi.org/10.1093/af/vfae025 Interpretive Summary: Society is concerned with ruminant livestock’s contribution to greenhouse gas emissions. Enteric methane (CH4) production represents the largest source of greenhouse gas emissions from beef and dairy production and has been identified as a source of emissions that must be reduced to meet commitments made by the U.S. government. Artificial intelligence has undergone recent advances, making it a powerful tool for research in numerous fields of study. Enteric CH4 mitigation is one field of study that artificial intelligence could support. Accordingly, researchers from ARS (Bushland), Iowa State University, and Kansas State University wrote this review paper to highlight the need for and potential applications of artificial intelligence for mitigating enteric CH4 emissions from ruminant livestock. In this review, we start with “Fermi”-calculations, which are simple calculations that demonstrate the benefit of a given practice – mitigating enteric CH4 in our instance. We then continue to discuss several applications of artificial intelligence to support enteric CH4 mitigation research. Technical Abstract: Enteric methane (CH4) emissions from livestock pose a significant challenge to greenhouse gas (GHG) reduction targets. • Generative AI and large language models may be able to predict molecules capable of inhibiting methanogenesis within the ruminant digestive system. Subsequent refinement through molecular simulations (protein–ligand docking and molecular dynamics) at physiological rumen pH allows for targeted compound selection. • Using a Graph Neural Network and the bovine metabolite database and milk composition database, compounds that have been experimentally shown to be anti-methanogenic can be identified using machine learning. Such biochemical representations can then be used to design new inhibitor molecules through generative AI. • While AI offers a promising avenue for effective CH4 mitigation, its implementation necessitates substantial computational resources and emphasizes the need for optimizing AI workflows to minimize energy expenditure while maximizing the discovery of effective mitigation strategies. |