GenVarLoader: an accelerated dataloader for applying deep learning to personalized genomics.
7 affiliations
- 1Bioinformatics and Systems Biology Program, University of California, San Diego, La Jolla, CA, 92093, United States.
- 2Integrative Biology Laboratory, Salk Institute for Biological Studies, La Jolla, CA, 92093, United States.
- 3Department of Computer Science and Engineering, University of California, San Diego, La Jolla, CA, 92093, United States.
- 4Herbert Wertheim School of Public Health and Longevity Science, University of California, San Diego, La Jolla, CA, 92093, United States.
- 5Integrative Biology Laboratory, Salk Institute for Biological Studies, La Jolla, CA, 92093, United States. [email protected].
- 6Bioinformatics and Systems Biology Program, University of California, San Diego, La Jolla, CA, 92093, United States. [email protected].
- 7Department of Medicine, Division of Genomics and Precision Medicine, University of California, San Diego, La Jolla, CA, 92093, United States. [email protected].
Abstract
Background
Deep learning models can predict the functional impact of genetic variants. However, training and applying these sequence models to personalized genome sequences are severely constrained by the computational burden of storing and reading massive datasets.Results
We developed GenVarLoader, an accelerated dataloader that creates personalized genomic sequences and functional tracks on-the-fly to overcome these bottlenecks. GenVarLoader stores this personalized genomic data in formats specifically optimized for machine learning. Our tool reads variants up to 26,000 times faster than a pre-subset BCF, loads personalized data up to 130 times faster than tuned FASTA and pyBigWig pipelines, and reduces storage approximately 2,000-fold compared to existing alternatives.Conclusions
By eliminating severe data loading bottlenecks, GenVarLoader ensures efficient model training and inference. It provides a scalable solution for integrating biobank-scale personalized genomic data with deep learning applications.
Related
- Applying Spatial Transcriptomics to the Human Heart: A Practical Guide for Translational Studies.
- Gut microbiota–cancer crosstalk: emerging molecular targets for personalized cancer treatment- a comprehensive review
- Feasibility and efficacy of personalized cardiorespiratory and strength exercise in Duchenne muscular dystrophy: a pilot study
- Sodium-dependent multivitamin transporter deficiency caused by novel SLC5A6 mutations: a clinical case and personalized mice model study
- Pharmacoepigenetics: from epigenetic regulation to personalized therapeutics
BibTeX
@article{laub2026genvarloader,
title = {{GenVarLoader: an accelerated dataloader for applying deep learning to personalized genomics.}},
author = {David Laub and Aaron Ho and Loukik Raina and Jeff Jaureguy and Adam Klie and Rany M Salem and Graham McVicker and Hannah Carter},
journal = {BMC genomics},
year = {2026},
month = jul,
doi = {10.1186/s12864-026-13225-5},
url = {https://europepmc.org/article/MED/42897450},
note = {CC BY},
}
RIS
TY - JOUR
TI - GenVarLoader: an accelerated dataloader for applying deep learning to personalized genomics.
AU - David Laub
AU - Aaron Ho
AU - Loukik Raina
AU - Jeff Jaureguy
AU - Adam Klie
AU - Rany M Salem
AU - Graham McVicker
AU - Hannah Carter
PY - 2026
DA - 2026/07/01
JO - BMC genomics
DO - 10.1186/s12864-026-13225-5
UR - https://europepmc.org/article/MED/42897450
AB - Background Deep learning models can predict the functional impact of genetic variants. However, training and applying these sequence models to personalized genome sequences are severely constrained by the computational burden of storing and reading massive datasets.Results We developed GenVarLoader, an accelerated dataloader that creates personalized genomic sequences and functional tracks on-the-fly to overcome these bottlenecks. GenVarLoader stores this personalized genomic data in formats specifically optimized for machine learning. Our tool reads variants up to 26,000 times faster than a pre-subset BCF, loads personalized data up to 130 times faster than tuned FASTA and pyBigWig pipelines, and reduces storage approximately 2,000-fold compared to existing alternatives.Conclusions By eliminating severe data loading bottlenecks, GenVarLoader ensures efficient model training and inference. It provides a scalable solution for integrating biobank-scale personalized genomic data with deep learning applications.
SN - 1471-2164
ER -