aiqbench analysis
HyCoSeq introduces a novel framework leveraging hyperbolic geometry and Lorentz convolutions to enhance genomic sequence representation learning without relying on massive pretraining datasets. By integrating weighted residual aggregation with bidirectional context modeling, the method achieves competitive performance against larger DNA language models while maintaining geometric consistency in local feature aggregation. This advancement highlights the potential of specialized mathematical inductive biases to improve efficiency and accuracy in bioinformatics applications.
HyCoSeq: Contextual Hyperbolic Representation Learning for Genomic Sequences
arXiv:2609.16925v1 Announce Type: new Abstract: Hyperbolic geometry provides a natural inductive bias for genomic representation learning, but existing hyperbolic genomic models primarily use Lorentz convolutions to learn local sequence representations, while their residual pathways do not directly aggregate full Lorentz representations. We propose HyCoSeq, a contextual hyperbolic representation learning framework for genomic sequences. HyCoSeq incorporates weighted Lorentzian residual aggregation into multi-curvature Lorentz encoding, allowing full Lorentz representations to participate directly in geometry-consistent local aggregation. It further introduces a bidirectional long short-term memory network that integrates information from both sequence directions to learn contextual relationships among local representations at different positions within a genomic sequence, thereby extending local hyperbolic convolutional encoding to sequence-level contextualized representations. Extensive experiments across diverse genomic tasks show that HyCoSeq outperforms existing hyperbolic baselines and, without large-scale genomic pretraining, achieves competitive performance against substantially larger pretrained DNA language models.
Abstract:Hyperbolic geometry provides a natural inductive bias for genomic representation learning, but existing hyperbolic genomic models primarily use Lorentz convolutions to learn local sequence representations, while their residual pathways do not directly aggregate full Lorentz representations. We propose HyCoSeq, a contextual hyperbolic representation learning framework for genomic sequences. HyCoSeq incorporates weighted Lorentzian residual aggregation into multi-curvature Lorentz encoding, allowing full Lorentz representations to participate directly in geometry-consistent local aggregation. It further introduces a bidirectional long short-term memory network that integrates information from both sequence directions to learn contextual relationships among local representations at different positions within a genomic sequence, thereby extending local hyperbolic convolutional encoding to sequence-level contextualized representations. Extensive experiments across diverse genomic tasks show that HyCoSeq outperforms existing hyperbolic baselines and, without large-scale genomic pretraining, achieves competitive performance against substantially larger pretrained DNA language models.
Machine Learning (cs.LG); Genomics (q-bio.GN); Machine Learning (stat.ML)
arXiv:2609.16925 [cs.LG]
(or arXiv:2609.16925v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2609.16925 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 236 words
Related coverage
SeqMaestro: From nucleotide sequences to biological hypotheses through interpretable machine learning
arXiv · Sep 16, 2026
PlantBGC: Transformer for Plant BGC Discovery via Label-Free Domain Adaptation and Weak Supervision
arXiv · Jul 31, 2026
RIBOSPAN: A Long-Context RNA Foundation Model for Versatile RNA Modeling
arXiv · Aug 25, 2026
AdaSurvMamba: Dynamic Fusion and Semantic Scanning for Multimodal Survival Analysis
arXiv · Jul 21, 2026