Through a unified two-stage alignment-and-generation pipeline, ProtBLIP2-SST integrates protein sequence and structural information, overcomes the rigidity of traditional GO-centric classification, generating open-ended captions that jointly describe molecular function, subcellular location, and homology context in one single output.
Abstract
Protein function prediction traditionally relies on structured gene ontology (GO) labels or multi-label classifiers. However, these labels or classifiers cannot flexibly describe molecular function, biological process, cellular component, and free-text functional narratives in a single output. In comparison, generation-based approaches offer an intuitive paradigm for flexible free-text protein annotation, with large language models (LLMs) as a representative method for protein-text modeling. Recent efforts on utilizing LLMs for protein semantic understanding and annotation generation have adopted sequence-only encoding or sequence-text contrastive alignment paradigms, yet without explicit consideration of three-dimensional structural information. To address these limitations in current protein function prediction methods, we present ProtBLIP2-SST, a two-stage framework built on the BLIP2 model architecture that bridges protein sequence, structure, and text for open-ended protein functional caption generation. Specifically, we first integrate sequence and structure information through SaProt, a protein language model (PLM) with a structure-aware vocabulary that fuses residue tokens with Foldseek-derived 3Di structural tokens. To empower the LLM to understand protein semantics, we employ a Q-Former (a querying transformer in BLIP2) with learnable query tokens as the cross-modal projector to align protein features from the frozen SaProt encoder and text features from a frozen BiomedBERT via protein–text contrasting, protein–text matching, and protein captioning objectives. After alignment, the protein features are linearly projected and prepended to the prompt embeddings of the LLM for protein captioning fine-tuning with LoRA. Trained on 441k protein–text pairs from Swiss-Prot with corresponding structures from the AlphaFold Database, our ProtBLIP2-SST outperforms sequence-only and sequence-text alignment baselines on protein captioning metrics, with ablation studies demonstrating the effectiveness of integrating structure with sequence information for improved protein understanding. Through a unified two-stage alignment-and-generation pipeline, ProtBLIP2-SST integrates protein sequence and structural information, overcomes the rigidity of traditional GO-centric classification, generating open-ended captions that jointly describe molecular function, subcellular location, and homology context in one single output.
Post-translational modifications (PTMs) and genetic variants regulate protein function, signalling, and disease, but their interpretation requires integration of sequence annotations with structural, interaction, and biophysical context. Although resources such as Scop3P, UniProt, the Protein Data Bank, and AlphaFold provide extensive annotations and structural information, integrating these data into reproducible structure-aware analyses still requires custom scripting and manual coordination between multiple independent tools. To address this challenge, we developed Scop3P-Toolkit, an open-source executable analytical environment for interactive analysis of PTMs, mutations, and proteomics-derived peptides in their structural context. The toolkit integrates protein annotation retrieval with structural mapping, residue interaction network analysis, comparative structural analysis, and residue-level biophysical profiling within a unified framework. Experimentally supported phosphosites, phosphopeptides, and phosphoproteomics evidence are provided for human proteins through Scop3P, with optional integration of curated UniProt PTM annotations. UniProt-derived PTMs, sequence features, and genetic variants are available for proteins from any species, extending the framework beyond the human phosphoproteome. Scop3P-Toolkit supports structure-centric analyses including interpretation of PTMs and disease-associated variants, analysis of residue interaction networks and their rewiring across alternative conformations, structural localisation of peptides, and exploration of protein–protein, protein–ligand, and host–pathogen interfaces. Interactive visualisation links sequence annotations, three-dimensional structures, residue interaction networks, and biophysical profiles, enabling coordinated exploration across multiple molecular representations. The toolkit is distributed as Jupyter notebooks, browser-based Voilà applications, and a Galaxy interactive tool, providing transparent, accessible, and reproducible workflows for both computational and experimental researchers. By integrating biological annotation resources into executable, structure-aware workflows, Scop3P-Toolkit enables reproducible interpretation of PTMs, mutations, and proteomics data.
Adrián Díaz, Natalia Tichshenko, Boris Depoortere et al.· bioRxiv· 0 citations
Applications to thioredoxins, visual opsins, and Tara Oceans environmental diatom cold-shock proteins show that PLMView can move from interpretable residue-level determinants in well-studied protein families to large-scale environmental functional discovery, linking molecular specialization to ecological distribution and transcriptional deployment across the global ocean.
WASP highlights how structural homology can systematically discover annotations missed by sequence-based approaches, predicting protein functions from AlphaFold structures using network-based structural homology and filling metabolic model gaps by mapping 75-100% of orphan reactions.
A neural network-based pipeline that integrates amino acid sequences with structural features is developed and provides a modular prototype for follow-up, more extensive protein modeling, including larger proteins and sequence of variable sizes.
Carl David Jasper Causin, M. Fyta· APL Machine Learning· 0 citations
De novo protein design is pivotal for revolutionizing protein engineering and advancing life sciences. Protein co-design aims to simultaneously create a novel protein sequence and structure with tailored functions, addressing the insufficient consistency between sequence and structure of two-stage design. Current AI-assisted protein co-design approaches primarily rely on protein sequence and structure information. However, they face two major challenges in limited integration of diverse biological knowledge and insufficient understanding of 'sequence-structure-function' relations, hindering the discovery of functional and diverse proteins in de novo design. To address these challenges, we propose a Protein sequence–structure–function Consistency Design model empowered by Natural Language function description, dubbed ProtcdNl, which expands the protein design space and ensures alignment with function-aware framework. Concretely, ProtcdNl contains two core components: i) the triple-coupled collaborative encoder, which achieves implicit alignment via joint latent space constraints, precisely maps natural language functional semantics to geometric protein motifs, and ii) a function-aware equivariant decoder, which endows the model with functional awareness while strictly maintaining the symmetry of molecular dynamics. Extensive experiments on our proposed ProtSSGT corpus demonstrate that ProtcdNl effectively mines the latent associations between functional semantics and protein geometry, achieving the design of novel, diverse proteins with high functional fidelity.
Ming Yang, Xin Zheng, Yi Li et al.· Proceedings of the 32nd ACM...· 0 citations
Multiple sequence alignment (MSA) Pairformer is presented, a protein language model that builds on AlphaFold2/3's bidirectional refinement between sequence and pairwise residue representations to accurately model the evolution of protein-protein interactions, despite training exclusively on individual chains.
Yo Akiyama, Zhidian Zhang, Olivia Tang et al.· Cell· 3 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
