Integrative/Hybrid Methods (IHM)

Structures of complex macromolecular assemblies are increasingly determined using integrative or hybrid methods (IHM), where a combination of complementary experimental and computational techniques is employed to model the structures. In addition to traditional structure determination methods such as X-ray crystallography (X-ray), NMR spectroscopy (NMR), and Electron Microscopy (3DEM), experimental techniques such as small angle scattering (SAS), atomic force microscopy (AFM), chemical crosslinking mass spectrometry (Crosslinking-MS), co-purification, F?rster resonance energy transfer (FRET), electron paramagnetic resonance (EPR), Hydrogen/Deuterium exchange (HDX), and various proteomics and bioinformatics approaches contribute to integrative structure determination. Spatial restraints derived from different kinds of experimental and computational methods are combined with known starting structures of molecular components to derive structures of macromolecular assemblies. Integrative modeling has been applied to determine the structures of complexes such as the nuclear pore complex and its sub-complexes, 16S rRNA complexed with methyltransferase A, human mitochondrial iron sulfur cluster core complex, the BBSome, nucleotide excision repair complex, ghrelin bound to its G protein-coupled receptor, and complex of RNF168-RING domain and the nucleosome.

Integrative structures can have heterogeneous model composition and can involve multi-scale or multi-state models, ordered states.

Multi-Scale Models

Multi-scaling supports representing a model as a collection of particles at different resolutions corresponding to atoms, single or multi-residue spherical beads, and 3D Gaussian objects. For example, a protein complex can be simultaneously described as a low-resolution volume representation of protein subunits as well as a well-resolved atomic representation of individual residues. Multi-scale representation allows for optimally encoding the model such that spatial restraints from input data can be accurately applied while retaining sufficient information to make the resulting models useful for further research.

Multi-State Models

A set of multiple states can be used to describe a system that exists in a mixture of multiple structural and/or compositional states that collectively satisfy the input information. For example, a sample of enzyme molecules in solution is structurally heterogeneous when it exists in an equilibrium between open and closed states; it is compositionally heterogeneous when it contains enzyme molecules both with and without a ligand.

Ordered States

The states in a multi-state model can be ordered in the form of a graph. This graph can be used to represent a model of a process such as an enzymatic reaction, a biochemical pathway, or a molecular dynamics trajectory.

Furthermore, integrative structures can be represented as a collection of models, where each one is consistent with given input information within an acceptable threshold, analogous to structures determined using NMR spectroscopy archived in the PDB. The variability among the models in the collection helps in assessing the uncertainty of modeling and the completeness of input data.

Why Integrative Structures Matter

Many essential biological assemblies—such as nuclear pore complexes, chromatin remodelers, viral capsids, and large protein-RNA machines — are too large, flexible, or heterogeneous to be fully resolved by a single method like X-ray crystallography, NMR, or cryo-EM. IHMs expand structural coverage of such large systems that are difficult to solve using a single method. By leveraging partial and lower-resolution datasets, IHMs broaden the range of macromolecular systems that can be structurally characterized. Furthermore, integrative modeling investigations can describe the kinetics and dynamics of biomolecular systems including their temporal evolution, energy landscapes, and motions that determine the transitions between states, thus providing additional insights into their function and interactions.

Supporting FAIR Principles

Integrative structures follow community-developed data standards based on the IHMCIF dictionary, which is a modular extension of the PDBx/mmCIF dictionary currently used by the wwPDB for archiving atomic structures of macromolecules. IHMCIF provides a flexible model representation to support multi-scale, multi-state, and ordered-state models. It also includes metadata definitions about modeling steps, software used, and sources of input data and spatial restraints. This promotes reproducibility and aligns with FAIR (Findable, Accessible, Interoperable, Reusable) data principles—crucial for modern, collaborative bioscience.

The IHMCIF dictionary is developed as a collaborative project that is distributed freely through a public GitHub repository.

Search and Access

RCSB.org provides multiple ways for users to discover and explore structures determined using Integrative/Hybrid Models (IHMs), which are now fully integrated into the website’s search and navigation. Integrative structures are part of unified search results.

Structure summary pages are provided for IHM entries, offering users an overview of key information about the representative model. Users can download the structure data in mmCIF format. Additionally, both the full validation report (PDF) and a summary validation report (PDF) are provided to offer comprehensive and high-level overviews of the model quality. Please note that PDB format files are not supported for IHM entries due to the complexity of these models, which cannot be adequately represented in the traditional PDB file format.

Note: All features may not be available for coarse-grained integrative structures. Furthermore, only data for representative models are available.

Keyword and Text Search

Integrative structures can be found using keyword-based searches in the Basic Search. For example:

• Searching for BBSome, Fibrin Clots, or human COP9 Signalsome
• Using PDB ID for a specific integrative structure, e.g., 8ZZE, 9A03, 9A0F

Using the Advanced Search Tool

To specifically find IHM entries:

• Go to the Advanced Search page
• Under the Structure Attributes section, chose Integrative/Hybrid Method Details
• Select from a list of options to search for integrative structures based on:
• Specific model features (e.g., multi-scale, multi-state, ordered)
• Different types of input experimental datasets (e.g., Crosslinking-MS data, 3DEM volume, NMR data etc.) and starting structural models (e.g., experimental model, de novo model)
• Accession codes of related datasets in external repositories (e.g., BMRB, EMDB, SASBDB, PRIDE etc.)

Using Refinements on Results Page

When you perform a general search (e.g., typing a protein name like “cohesin”), you can filter the results to only show IHMs:

• On the left-hand side of the search results page, find the Structure Determination Methodology category
• Check the box for integrative
• Additional filters like Release Date or Scientific Name of Source Organism can further narrow results

Distinct Visual Markers

IHM icon shown in structure summary and results pages

Programmatic Access

You can retrieve IHM entries using the RCSB Search and Data APIs. Full API documentation is available at search.rcsb.org and data.rcsb.org.

Examples

• All integrative structures in the archive: search example
• Integrative structures released in 2024: search example
• Integrative structures of human proteins that use crosslinking mass spectrometry as part of the modeling process: search example
• Integrative structures of large macromolecular assemblies (more than 10 polymer entities): search example
• Integrative structures generated using specific modeling software: search example

References

• Vallat B et al., J Mol Biol. 2025; 168963, doi: 10.1016/j.jmb.2025.168963
• Vallat B et al., J Mol Biol. 2024; 436(17): 168546. doi: 10.1016/j.jmb.2024.168546
• Vallat B et al., Acta Cryst. 2021; D77: 1486-1496, doi: 10.1107/S2059798321010871
• Sali A. J Biol Chem. 2021; 296: 100743. doi: 10.1016/j.jbc.2021.100743
• Berman HM et al., Structure. 2019 Dec 3; 27(12): 1745-1759. doi: 10.1016/j.str.2019.11.002
• Vallat B et al., J Biomol NMR. 2019 Jul; 73(6-7): 385-398. doi: 10.1007/s10858-019-00264-2
• Vallat B et al., Structure. 2018 Jun; 26(6):894-904, doi: 10.1016/j.str.2018.03.011
• Burley SK et al., Structure. 2017 Sep; 25(9):1317-1318. doi: 10.1016/j.str.2017.08.001
• Sali A et al., Structure. 2015 Jul; 23(7):1156-1167. doi: 10.1016/j.str.2015.05.013