ASOFormer: a Transformer-based model for predicting antisense oligonucleotide efficacy to support therapeutic candidate prioritization
- 1Mayo Clinic
Abstract
Antisense oligonucleotides (ASOs) represent a promising therapeutic modality for RNA-based disease mechanisms, but identifying high-efficacy candidates from large sequence spaces remains a major bottleneck in drug development. Here, we present ASOFormer, a novel Transformer-based neural network that predicts the inhibition efficiency of RNase H-mediated ASOs from sequence, predicted secondary structure, and chemical modification features. ASOFormer was trained on a knockdown efficacy dataset spanning over 170,000 ASO-target pairs compiled largely from published patents. Ablation analysis confirmed that the addition of chemical modification features to the primary sequence drove the largest performance gains, with predicted secondary structure providing complementary benefit when combined with modification information. Further analysis highlighted the importance of the wings of gapmer ASOs and other features. As independent validation, ASOFormer was applied to ASOs targeting seven genes that were not in the training dataset gene list. For four out of the seven genes, the ground truth ASO inhibition efficiency was obtained from a public dataset, whereas the others were measured in-house via qPCR. Compared to two published state-of-the-art methods, ASOFormer achieved the best overall accuracy and prioritized the strongest inhibitors. Crucially, only ASOFormer consistently exceeded random expectation for top-candidate recovery across all seven test genes.
Related
- DEPICT enables sample-centric therapeutic prioritization across evolving and spatially heterogeneous tumor states
- Biopsy-Based Molecular Signatures Predicting Therapeutic Response in Ulcerative Colitis
- Phylogenetic Analysis of cpn60 UT region reveal oligonucleotide regions for the detection of prominent Enterobacteriaceae pathogens
- Spatial oligonucleotide immuno-reprogramming model (SOIRM): a testable framework linking bone marrow resistance states and oligonucleotide therapeutics in multiple myeloma
- A compact three-way junction oligonucleotide probe for widefield and super-resolution imaging of RNAs and proteins
BibTeX
@misc{dove2026asoformer,
title = {{ASOFormer: a Transformer-based model for predicting antisense oligonucleotide efficacy to support therapeutic candidate prioritization}},
author = {N. C. Dove and Y. Min and O. Is and N. Ertekin-Taner and D. P. Wickland and X. Wang},
howpublished = {bioRxiv},
year = {2026},
month = sep,
doi = {10.64898/2026.09.18.752788},
url = {https://www.biorxiv.org/content/10.64898/2026.09.18.752788v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - ASOFormer: a Transformer-based model for predicting antisense oligonucleotide efficacy to support therapeutic candidate prioritization
AU - N. C. Dove
AU - Y. Min
AU - O. Is
AU - N. Ertekin-Taner
AU - D. P. Wickland
AU - X. Wang
PY - 2026
DA - 2026/09/18
PB - bioRxiv
DO - 10.64898/2026.09.18.752788
UR - https://www.biorxiv.org/content/10.64898/2026.09.18.752788v1
AB - Antisense oligonucleotides (ASOs) represent a promising therapeutic modality for RNA-based disease mechanisms, but identifying high-efficacy candidates from large sequence spaces remains a major bottleneck in drug development. Here, we present ASOFormer, a novel Transformer-based neural network that predicts the inhibition efficiency of RNase H-mediated ASOs from sequence, predicted secondary structure, and chemical modification features. ASOFormer was trained on a knockdown efficacy dataset spanning over 170,000 ASO-target pairs compiled largely from published patents. Ablation analysis confirmed that the addition of chemical modification features to the primary sequence drove the largest performance gains, with predicted secondary structure providing complementary benefit when combined with modification information. Further analysis highlighted the importance of the wings of gapmer ASOs and other features. As independent validation, ASOFormer was applied to ASOs targeting seven genes that were not in the training dataset gene list. For four out of the seven genes, the ground truth ASO inhibition efficiency was obtained from a public dataset, whereas the others were measured in-house via qPCR. Compared to two published state-of-the-art methods, ASOFormer achieved the best overall accuracy and prioritized the strongest inhibitors. Crucially, only ASOFormer consistently exceeded random expectation for top-candidate recovery across all seven test genes.
ER -