HD-Prot: New Protein Language Model Integrates Sequence and Structure Data Using Continuous Tokens
Researchers have proposed HD-Prot, a hybrid diffusion protein language model that integrates continuous structural information directly into a sequence-based protein language model without discretizing structural data. Unlike existing approaches that convert protein structures into discrete tokens—losing fine-grained detail in the process—HD-Prot uses a continuous diffusion head atop a discrete language model to handle both modalities simultaneously. The method achieves competitive performance across multiple protein design and prediction tasks while requiring less than one-tenth the computational budget of comparable state-of-the-art models.
HD-Prot (Hybrid Diffusion Protein Language Model) addresses a longstanding challenge in computational protein science: how to meaningfully incorporate three-dimensional structural information into protein language models (pLMs) that are primarily designed for discrete sequence data. Current multimodal pLMs typically discretize protein structures via vector quantization, a step that sacrifices fine-grained structural detail. HD-Prot avoids this by embedding a continuous-valued diffusion head on top of a discrete pLM, enabling the model to process both discrete sequence tokens and continuous structure latents within a unified absorbing diffusion framework. The model estimates per-token distributions using categorical prediction for sequences and continuous diffusion for structures, capturing cross-modal dependencies between the two. Benchmarks show HD-Prot performs competitively in unconditional sequence-structure co-generation, motif-scaffolding, protein structure prediction, and inverse folding—key tasks in protein engineering and drug design. Notably, these results were achieved with a modality-extension fine-tuning budget under one-tenth that of leading multimodal pLMs, suggesting strong computational efficiency. The work is accepted to KDD 2026 and the preprint represents the extended version of that paper.
What's missing
The paper does not report results on experimental wet-lab validation of designed proteins, so it remains unclear how computationally generated sequences and structures translate to real-world biochemical function. Benchmark comparisons are limited to computational metrics.
What different sources said
- arXiv cs.AICenter
HD-Prot: A Protein Language Model for Joint Sequence-Structure Modeling with Continuous Structure Tokens
Related
Gut Bacteria Enzyme Found to Break Down Heat-Processed Food Compounds, Producing Novel Biogenic Amines
Researchers have discovered that an enzyme in common gut bacteria can degrade N-epsilon-carboxymethyllysine (CML), a compound formed during thermal food processing, producing previously unknown biogenic amines. The enzyme, ornithine decarboxylase SpeC from enterobacteria, acts on CML and related modified lysine derivatives through a low-level 'underground' catalytic activity. This finding suggests a previously unrecognized communication axis between thermally processed dietary compounds and gut microbial physiology, with potential implications for host health.
Full-Length Gene Sequencing Reveals Two Distinct Bacterial Communities in Black-Legged Ticks Expanding Into Canada
Researchers used Oxford Nanopore full-length 16S rRNA gene sequencing to characterize the microbiome of Ixodes scapularis black-legged ticks collected in Nova Scotia, Canada, distinguishing between tick-adapted bacteria and environmentally acquired bacteria. The study comes as I. scapularis — the primary vector of Lyme disease — is rapidly expanding northward into Canada due to climate change. The findings suggest that environmentally derived bacteria in tick microbiomes are not mere contamination, which has implications for how tick microbiome data is collected and interpreted across surveillance studies.
Study Identifies Metabolic Link Between Cell Envelope Stress and Biofilm Formation in Bacteria
Researchers have discovered that the metabolite acetyl-CoA directly inhibits enzymes that degrade the bacterial signaling molecule c-di-GMP, connecting cell envelope biosynthesis stress to biofilm formation in Pseudomonas aeruginosa. The study found that sub-inhibitory concentrations of antibiotics targeting early peptidoglycan biosynthesis — but not other antibiotic classes — elevate c-di-GMP levels by reducing phosphodiesterase activity, with acetyl-CoA competing for the enzyme active site. Because the relevant enzyme domain is broadly conserved across bacterial species, this checkpoint mechanism may be widespread and could have implications for understanding antibiotic-induced biofilm responses.