Related Experiment Video
Updated: May 16, 2025
Application of Long-term cultured Interferon-γ Enzyme-linked Immunospot Assay for Assessing Effector and Memory T Cell Responses in Cattle
Published on: July 11, 2015
Using Supervised Machine Learning Algorithms to Predict Bovine Leukemia Virus Seropositivity in Florida Beef Cattle:
Ameer A Megahed1,2, Y Reddy Bommineni1,3, Michael Short3
1Department of Large Animal Clinical Sciences, College of Veterinary Medicine, University of Florida, Gainesville, Florida, USA.
Background:
Bovine leukemia virus (BLV) infection in beef cattle has received less attention than in dairy herds, despite its potential impact on the beef industry.
Objectives:
To compare six different supervised machine-learning (SML) algorithms used to identify the most important risk factors for predicting BLV seropositivity in beef cattle in Florida.
Animals:
Retrospective study. We used a dataset of 1511 blood sample records from the Bronson Animal Disease Diagnostic Laboratory, Florida Department of Agriculture & Consumer Services, submitted for BLV antibody testing from 2012 to 2022.
Methods:
Logistic regression (LR), decision tree (DT), gradient boosting (GB), random forest (RF), neural network (NN), and support vector machine (SVM) were used.
Results:
Of the submitted samples, 11.6% were positive for BLV. The RF model best predicted BLV infection with an area under the receiver operating characteristic curve (AUROC) of 0.98, with a misclassification rate of 0.06. The DT model showed comparable performance to RF (AUROC, 0.94; misclassification rate, 0.06). However, the NN model had the poorest performance. The RF model showed that BLV seropositivity can be best predicted by testing beef cows during the dry season, which mostly coincides with the pre-calving processing and calving seasons, particularly for cattle raised in southern Florida.
Conclusions:
The RF model shows promise for predicting BLV seropositivity in beef cattle. Key predictive risk factors include the dry season months coinciding with pre-calving and calving seasons and geographic location. These findings could help develop predictive tools for effective screening for BLV infection and targeted interventions.
