Machine Learning Identification of Low Methane Emitting Dairy Cattle for Selective Breeding

Stephen Ross, Huiru Zheng, Haiying Wang, Tianhai Yan, Masoud Shirali

Research output: Contribution to conferencePosterpeer-review

11 Downloads (Pure)


The rapid evolution of humanity has led to a drastic increase in Green House Gas (GHG) levels in the atmosphere. This has caused an unprecedented rise in the Earth’s surface temperature, resulting in Global Warming and Climate Instability. 30% of the current rise in global temperature is due to Methane (CH4). However, approximately 60% of global CH4 emissions are man-made.
Agriculture is the primary source of anthropogenic CH4 emissions, from which 40% originates, primarily generated by Livestock. Cattle are the most prolific contributor of CH4 amongst all Livestock, responsible for 77% of emissions.
Despite the current situation, the human population continues to grow, and is expected to reach 9.7 billion by 2050, placing even further pressure on Agricultural production systems. Therefore, Cattle CH4 emissions must be harnessed, so that the expected population growth can be catered for, without further damaging the Earth’s already strained climate.
Selective Breeding of low CH4 emitting dairy cattle (DC) is a cutting-edge approach in the mitigation of Agricultural CH4 emissions. Several studies have shown that CH4 emissions in DC have a genetic component which is heritable. Therefore, Selective Breeding of low CH4 emitting DC has a permanent, compounding effect over time, as future generations inherit the low CH4 emission characteristics of their more efficient ancestors.
Machine Learning (ML) models are a groundbreaking option for the identification of low CH4 emitting DC. Their incredible ability to facilitate cross talk between a diverse range of features, makes them ideal for modelling the intricate relationships between the biological, environmental, and genetic factors behind CH4 production.
Therefore, this PhD project aims to develop a ubiquitous ML framework which can incorporate complex biological, environmental and genetic factors for the accurate prediction of DC CH4 emissions so that low CH4 emitting DC can be identified for Selective Breeding.
Original languageEnglish
Publication statusPublished online - May 2023
EventVirtual Institute of Bioinformatics and Evolution - Queens University, Belfast, United Kingdom
Duration: 12 May 202312 May 2023


ConferenceVirtual Institute of Bioinformatics and Evolution
Country/TerritoryUnited Kingdom


  • Dairy Cattle
  • Methane
  • Machine Learning
  • Selective Breeding


Dive into the research topics of 'Machine Learning Identification of Low Methane Emitting Dairy Cattle for Selective Breeding'. Together they form a unique fingerprint.

Cite this