Evaluating imputation methods for determining North American honey bee (Apis mellifera) ancestry from low-pass whole-genome sequencing data
Abstract
Honey bee (Apis mellifera) movement is tightly regulated throughout North America to prevent disease spread and the introduction of unwanted subspecies, such as the ‘scutellatized’ bee (also known as “Africanized,” a mix of A. m. scutellata and managed European bees). In this study, we evaluated the effectiveness of low-coverage whole-genome sequencing (3X) with genotype imputation for ancestry estimation and phenotypic prediction in 289 honey bee samples from the USA and Caribbean. Imputation showed high accuracy (mean r2 = 85%, median r2 = 90%) and concordance (r > 95%) with full-depth sequencing, indicating this approach can reliably detect common genetic variation in North American populations. We tested this method’s utility in a key regulatory application: detecting scutellatized ancestry. Low-pass sequence data allowed both genome-wide ancestry estimation and mitochondrial haplotype reconstruction, the latter showing near-perfect agreement with a widely used PCR test. We found that mitochondrial haplotypes were poor predictors of genomic ancestry, and genomic ancestry a better predictor of colony phenotypes. This questions the reliability of mitochondrial-DNA-based detection methods alone in the face of complicated systems of hybridization. Our findings demonstrate that low-coverage sequencing with imputation is a powerful tool for enabling whole genome analyses in honey bees—including ancestry assignment and variant detection—despite a small loss in accuracy. It will likely find application in regulatory screening, design breeding schemes, and population studies.
Related
- Biocontrol potential of hive-associated Bacillus species against American Foulbrood disease in honeybees (Apis mellifera L.).
- RNA sequencing of antennae and comparative analysis of key gene families in Apis mellifera workers exhibiting drastic differences in propolis yield
- Evaluating Clustering and Imputation Approaches for Multivariate Longitudinal Clinical Data Using a Real-World eCDF-Copula Resampling Framework
- Alert-List Concordance for Evaluating Imputation Strategies in ICU Sepsis Prediction
- Evaluating and improving AI agents for sequencing protocol understanding
BibTeX
@article{ryals2026evaluating,
title = {{Evaluating imputation methods for determining North American honey bee (Apis mellifera) ancestry from low-pass whole-genome sequencing data}},
author = {Dylan K. Ryals and Mulusew Fikere and Madeline H. Carpenter and J Krispn Given and James C. Fulton and Matthew R. Moore and Ethel M. Villalobos and Robyn Underwood and Myra Dickey and Andrew D. Sweet and Julie M. Allen and Juliana Rangel and Margarita M. López-Uribe and Luiz F. Brito and Andrew H. Paterson and Brock A. Harpur},
journal = {Apidologie},
year = {2026},
month = oct,
doi = {10.1007/s13592-026-01311-2},
url = {https://hal.science/hal-05793438v1},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Evaluating imputation methods for determining North American honey bee (Apis mellifera) ancestry from low-pass whole-genome sequencing data
AU - Dylan K. Ryals
AU - Mulusew Fikere
AU - Madeline H. Carpenter
AU - J Krispn Given
AU - James C. Fulton
AU - Matthew R. Moore
AU - Ethel M. Villalobos
AU - Robyn Underwood
AU - Myra Dickey
AU - Andrew D. Sweet
AU - Julie M. Allen
AU - Juliana Rangel
AU - Margarita M. López-Uribe
AU - Luiz F. Brito
AU - Andrew H. Paterson
AU - Brock A. Harpur
PY - 2026
DA - 2026/10/12
JO - Apidologie
DO - 10.1007/s13592-026-01311-2
UR - https://hal.science/hal-05793438v1
AB - Honey bee (Apis mellifera) movement is tightly regulated throughout North America to prevent disease spread and the introduction of unwanted subspecies, such as the ‘scutellatized’ bee (also known as “Africanized,” a mix of A. m. scutellata and managed European bees). In this study, we evaluated the effectiveness of low-coverage whole-genome sequencing (3X) with genotype imputation for ancestry estimation and phenotypic prediction in 289 honey bee samples from the USA and Caribbean. Imputation showed high accuracy (mean r2 = 85%, median r2 = 90%) and concordance (r > 95%) with full-depth sequencing, indicating this approach can reliably detect common genetic variation in North American populations. We tested this method’s utility in a key regulatory application: detecting scutellatized ancestry. Low-pass sequence data allowed both genome-wide ancestry estimation and mitochondrial haplotype reconstruction, the latter showing near-perfect agreement with a widely used PCR test. We found that mitochondrial haplotypes were poor predictors of genomic ancestry, and genomic ancestry a better predictor of colony phenotypes. This questions the reliability of mitochondrial-DNA-based detection methods alone in the face of complicated systems of hybridization. Our findings demonstrate that low-coverage sequencing with imputation is a powerful tool for enabling whole genome analyses in honey bees—including ancestry assignment and variant detection—despite a small loss in accuracy. It will likely find application in regulatory screening, design breeding schemes, and population studies.
SN - 0044-8435
SN - 1297-9678
ER -