# Protein Structure Tokenization via Geometric Byte Pair Encoding

arXiv: https://arxiv.org/abs/2511.11758 · HTML: https://mlsanity.com/paper/2511.11758/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 9 of its 58 references that are in the map.
- First posted on arXiv 13 Nov 2025.

## Builds on (most cited first, 9 of 9)

- [Neural Discrete Representation Learning](https://mlsanity.com/paper/1711.00937/) (2017)
- [Diffusion probabilistic modeling of protein backbones in 3D for the motif-scaffolding problem](https://mlsanity.com/paper/2206.04119/) (2022)
- [Evaluating Protein Transfer Learning with TAPE](https://mlsanity.com/paper/1906.08230/) (2019)
- [HyenaDNA: Long-Range Genomic Sequence Modeling at Single Nucleotide Resolution](https://mlsanity.com/paper/2306.15794/) (2023)
- [DNABERT-2: Efficient Foundation Model and Benchmark For Multi-Species Genome](https://mlsanity.com/paper/2306.15006/) (2023)
- [Protein structure generation via folding diffusion](https://mlsanity.com/paper/2209.15611/) (2022)
- [Codebook Transfer with Part-of-Speech for Vector-Quantized Image Modeling](https://mlsanity.com/paper/2403.10071/) (2024)
- [Protein Structure Tokenization: Benchmarking and New Recipe](https://mlsanity.com/paper/2503.00089/) (2025)
- [Linguistic Laws Meet Protein Sequences: A Comparative Analysis of Subword Tokenization Methods](https://mlsanity.com/paper/2411.17669/) (2024)

## Built on by (most cited first, 4 of 4)

- [ENSEMBITS: an alphabet of protein conformational ensembles](https://mlsanity.com/paper/2605.13789/) (2026)
- [PUFFIN: Protein Unit Discovery with Functional Supervision](https://mlsanity.com/paper/2604.14796/) (2026)
- [LEMON-ZEST: Evolution-Informed Tokenization for Efficient Protein Language Modeling](https://mlsanity.com/paper/2609.37675/) (2026)
- [ProTok: Unified Protein Tokenization via Modular Multi-Codebook Quantization](https://mlsanity.com/paper/2610.18481/) (2026)

## Abstract (arXiv)

Protein structure is central to biological function, and enabling multimodal protein models requires joint reasoning over sequence, structure, and function. A key barrier is the lack of principled protein structure tokenizers (PSTs): existing approaches fix token size or rely on continuous vector codebooks, limiting interpretability, multi-scale control, and transfer across architectures. We introduce GeoBPE, a geometry-grounded PST that transforms continuous, noisy, multi-scale backbone conformations into discrete ``sentences'' of geometry while enforcing global constraints. Analogous to byte-pair encoding, GeoBPE generates a hierarchical vocabulary of geometric primitives by iteratively (i) clustering Geo-Pair occurrences with k-medoids to yield a resolution-controllable vocabulary; (ii) quantizing each Geo-Pair to its closest medoid prototype; and (iii) reducing drift through differentiable inverse kinematics that optimizes boundary glue angles under an SE(3) end-frame loss. GeoBPE offers compression (>10x reduction in bits-per-residue at similar distortion rate), data efficiency (>10x less training data), and generalization (maintains test/train distortion ratio of 1.0-1.1). It is architecture-agnostic: (a) its hierarchical vocabulary provides a strong inductive bias for coarsening residue-level embeddings from large PLMs into motif- and protein-level representations, consistently outperforming leading PSTs across 12 tasks and 24 test splits; (b) paired with a transformer, GeoBPE supports unconditional backbone generation via language modeling; and (c) tokens align with CATH functional families and support expert-interpretable case studies, offering functional meaning absent in prior PSTs. Code is available at https://github.com/shiningsunnyday/PT-BPE/.