Table 5.
Examples of ML studies and the potential of digital twin to develop the results outcomes.
| Animal species | Study | Studied variable | Marker | References | DT upgrading potential |
|---|---|---|---|---|---|
| Dairy cattle | Developing an artificial neural network for the early prediction of subclinical ketosis during lactation. | Subclinical Ketosis | 50,025 and 10,005 SNPs | (185) | Predict the effect on lactation performance and yield, and forecast culled animals. |
| Beef cattle (Nellore) | Utilizing machine learning to find small subsets of biologically significant genes for classifying animals into High Feed Efficiency and Low Feed Efficiency categories. | Feed efficiency | 16,423 genes | (186) | Rearranging this subset or investigating new gene subsets to predict their effect on the traits. |
| Goat | Efficiently oversee the health and welfare of their goats, thereby enhancing living circumstances and augmenting dairy output. | Animal behavior | Goat activity | (187) | Mimic population behavior changes and track their effect on herd performance. |
| Dairy cattle | Image processing algorithms and the YOLOv8 model facilitate the real-time, non-invasive monitoring of feeding periods. | Feed utilization efficiency | Feeding pattern | (188) | Simulate different feeding strategies and forecast productivity and economics. |
| Cattle and Buffalo | Forecasting lumpy skin disease infection | LSD occurrence | Meteorological and geological attributes | (189) | Stimulate assessment of infection virulence and its effects on productivity, and estimate potential economic losses in specific regions. |
| (Sheep) (Harnai) | Predicting live weight at the post-weaning period. | Growth performance | Body biometric parameters and sex factor. | (190) | Predicting the Economics of the Production |
| Pigs | Infection prediction in swine populations, both seven and 30 days in advance | Infection Outbreak | Nearby farm density, historical test rates, piglet inventory, feed consumption during gestation, and wind speed and direction. | (95) | Simulate the evaluation for disease prevention and mitigation strategies. |
| Cattle | 16S rRNA sequencing and machine learning methodologies identified a dozen species as taxonomic indicators for distinguishing infection. | The Mycobacterium avium disease state. | Fecal microbiota | (191) | Investigate the effects of microbial dysbiosis on pathogenic microbes. The Gut-host interaction |
| Chicken | Investigate antimicrobial resistance profiles across multiple chicken farms and abattoirs. | Antimicrobial resistance genes | E. coli | (192) | Forecast the intensity of ARGs under different environmental conditions and correlate ARGs with different microbiomes. |
| In-vitro fermentation | Prediction of Methane Production from in vitro Ruminal Fermentation | Methane | Volatile fatty acids | (193) | Introducing a cross-species Virtual rumen fermentation model |
| Water-Deer Water Buffalo Sheep Buffalo |
Combined network analysis and interpretable machine learning reveal | Environmental adaptability | Microbial genomes | (194) | Predicting which pathogen species are most likely to emerge in the future |
| Wild-Livestock | ML demonstrates one approach to planning for and preventing disease emergence in livestock. | Pathogen-host associations at the wildlife–livestock interface | Bacterial association | (195) |