Aug 2026· International Journal of Molecular Sciences· Vol 27· 0 citations· 51 references
Medicine
TL;DR
This review focuses on coordinate- and residue-frame-based diffusion approaches for generating protein structures, paying particular attention to geometric equivariance, conditioning strategies, all-atom modelling and interaction-aware design.
Abstract
Although deep learning has transformed protein structure prediction, the controlled generation of functional and experimentally tractable protein structures remains a major challenge in structural bioinformatics. Diffusion models offer a versatile approach to generating protein backbones, motif-conditioned scaffolds, all-atom structures and biomolecular interaction geometries, while accommodating explicit structural and functional constraints. This review focuses on coordinate- and residue-frame-based diffusion approaches for generating protein structures, paying particular attention to geometric equivariance, conditioning strategies, all-atom modelling and interaction-aware design. We compare representative methods derived from RoseTTAFold, frame-diffusion architectures, and oriented-residue-cloud representations according to their molecular representation, generative objective, and validation strategy. We examine the criteria used to evaluate generated proteins, such as stereochemical quality, structural consistency, designability, novelty, diversity, computational efficiency, and experimental performance. Particular attention is given to the distinction between learned structural distributions and condition-dependent thermodynamic ensembles. Future progress will depend on the integration of generative models with molecular mechanics, conformational sampling, uncertainty estimation, free-energy methods, and experimental design–build–test–learn cycles. Within this framework, diffusion models offer candidate generation and constraint satisfaction capabilities within broader protein engineering workflows.
Protein engineering relies heavily on computational characterization of constrained protein fitness landscapes, in which only a limited fraction of sequence space corresponds to stable and functional biomolecules. Advances in structural biology and machine learning are progressively shifting protein design strategies from empirical optimization toward multidimensional evaluation of sequence-structure-function relationships. This review examines current computational strategies for exploring these landscapes, including sequence-derived evolutionary descriptors, structural fitness assessment, energetic evaluation, and integrated multi-parameter scoring. Recent developments in protein language models, deep-learning-based structure prediction, generative protein design, and consensus scoring approaches support large-scale exploration of biologically accessible sequence space. Negative-design constraints, including aggregation propensity, intrinsic disorder, and developability are important in prioritizing experimentally tractable protein candidates. Finally, the integration of computational prediction with iterative experimental validation is discussed as a central framework for rational protein engineering. By framing structure prediction, sequence representation learning, and generative design as complementary strategies for navigating a single constrained fitness landscape, this review highlights integrated, multidimensional scoring and negative-design filtering as the critical link between computational candidate generation and experimentally tractable protein design.
OrgNet+, a conformational ensemble-aware and orientation-gnostic framework that explicitly incorporates protein structure flexibility during training, is introduced, which substantially reduces intra-ensemble prediction variance while simultaneously improving predictive accuracy.
A. Sarycheva, Aleksandr Shumilov, Petr Popov· Bioinformatics· 0 citations
We model protein structural stability as a continuous scalar quantity defined over molecular geometry, referred to as the support field. Instead of treating stability as a discrete residue annotation or an empirical score, this representation characterizes protein folds through the combined effects of geometric organization, topological persistence, and local density. Based on this idea, we introduce Support Field Neural Representation Learning (SF-NRL), a topology-guided approach that integrates persistent homology(PH), spatial density estimation, and geometric deep learning to infer residue-wise support directly from protein structures. Persistent topological features are incorporated as structural constraints that modulate local support values across the fold, enabling a continuous description of structural reliability. Across diverse protein families, the inferred support field shows consistent agreement with independent indicators of structural stability and highlights low-support regions associated with conformational flexibility and weak structural integration. By embedding protein structures into a continuous stability landscape, SF-NRL provides an interpretable representation that complements structure prediction models and facilitates systematic identification of structural cores, flexible regions, and functionally relevant motifs. These results demonstrate that topology-informed field representations offer a generalizable and practically useful approach for analyzing protein stability and fold organization.
Accurate protein structure prediction is fundamental to structural biology because protein structure underlies molecular function and provides a basis for mechanistic interpretation. Recent advances in deep learning have transformed the field from multiple sequence alignment (MSA)-driven monomer folding into broader frameworks capable of modeling protein complexes and increasingly heterogeneous molecular systems. Existing reviews have summarized this progress from the perspectives of representative models, application domains, and protein design. Building on these efforts, this review focuses on the methodological evolution of the field itself. It examines recent developments through three closely related dimensions: representations and data, architectures and learning strategies, and confidence and evaluation. Within this perspective, the field is organized into four methodological phases and three cross-cutting transitions: from explicit evolutionary coupling features and early contact prediction to learned sequence representations in AlphaFold2, RoseTTAFold, and ESMFold; from protein-only monomer folding to increasingly integrated modeling of heterogeneous molecular systems in AlphaFold-Multimer, RoseTTAFoldNA, and AlphaFold3; and, more recently, from prediction-oriented structure inference to design-oriented generative modeling in RFdiffusion and related frameworks. This framework provides a clearer understanding of how methodological shifts have shaped the capabilities, limitations, and practical roles of recent models.
Wengan He, Yongsheng Luo, Lihong Jiang et al.· 0 citations
The quantitative prediction of biomolecular recognition is crucial to molecular science. The challenge is not merely structural determination but the prediction of (thermo)dynamic and kinetic observables arising from high-dimensional molecular ensembles, such as free energies, conformational distributions, and rate processes across different conditions. As the field shifts from structure-centric to ensemble-based descriptions, two complementary modeling strategies have matured: explicit energy-based approaches grounded in statistical mechanics and data-driven models that learn statistical representations of molecular configurations from large data sets. Physics-based methods, including molecular dynamics and free energy perturbation, estimate observables by sampling (Boltzmann-distributed) configurations under approximate molecular Hamiltonians, thereby providing mechanistic interpretability and thermodynamic consistency, albeit at non-negligible computational cost and with inherent force field limitations. In contrast, modern machine learning approaches rapidly generate structures and propose conformational ensembles without explicit thermodynamic weighting, by learning statistical patterns in structural and bioactivity data. While these methods often achieve high predictive performance, they do not inherently enforce thermodynamic consistency due to the lack of an explicit connection to a partition function and thus may produce configurations that are not physically realizable. We argue that, since physics-based simulations and machine learning provide complementary approximations to the underlying probability distribution associated with biomolecular recognition events, and they excel respectively in consistency with free-energy landscapes and state populations and in predictive accuracy, the central challenge for the coming decade will be integrating them into hybrid frameworks that are scalable and transferable.
R. Khalil, Elena Frasnetti, Han Kurt et al.· Journal of Physical Chemistr...· 0 citations
Structure-prediction networks built on co-evolutionary statistics have transformed protein-based drug discovery, yet their accuracy does not extend to peptide therapeutics--an increasingly important modality defined by non-canonical residues, macrocyclization, and complex topologies. We introduce Vilya-2, a diffusion transformer that extends the all-atom representation of Vilya-1 from modeling individual molecules to modeling their interactions with protein targets. This all-atom representation enables transfer learning between different molecular types, and delivers highly accurate structural modeling of peptides across sizes, classes, and compositions bound to therapeutically relevant targets. By generating diverse structural ensembles and ranking them with calibrated confidence, Vilya-2 recovers 59.1% of peptide interfaces to sub-2 {\AA} backbone RMSD, far exceeding the performance of a representative co-folding model even when that model is given the bound receptor as a template. In addition, Vilya-2 is state-of-the-art at small-molecule docking, and generalizes to novel protein-small molecule complexes unlike those seen in training. It also generalizes to modeling molecular conformations of diverse macrocycles and disulfide-stapled miniproteins several-fold larger than any molecule seen in training. Finally, Vilya-2 can be used as a foundation model, and fine-tuned to enrich for active compounds in hit-to-lead campaigns. By unifying predictive accuracy with broad generalizability across chemical space, Vilya-2 is the structure-prediction oracle that de novo peptide design pipelines require--establishing the all-atom approach as a general foundation for the design and evaluation of de novo peptide therapeutics.
Vilya Research Pascal Sturmfels, Naozumi Hiranuma, M. Salem et al.· 0 citations