Hycoseq improves dna sequence analysis using hyperbolic geometry
HyCoSeq: Contextual Hyperbolic Representation Learning for Genomic Sequences
Machine Learning
Summary
DNA sequences contain complex patterns that are important to understand for many biology and medicine tasks. The authors developed HyCoSeq, a new way to represent DNA sequences using a special kind of geometry called hyperbolic geometry, which helps capture these patterns better. HyCoSeq uses advanced techniques to look at pieces of DNA in both directions and combine them into a detailed representation. This leads to better performance on various DNA analysis tasks compared to existing models, even without huge amounts of pretraining data.
What this means in practice
- •For genomic data scientists: Enhance DNA sequence pattern recognition in computational pipelines with improved hyperbolic representation approaches for better task accuracy.
- •For bioinformatics software developers: Integrate HyCoSeq to build DNA analysis tools that require efficient and accurate sequence embeddings without relying on large pretrained models.
Authors
Chenhao Zeng, Zhibin Pu, Shufei Ge
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.