- Atlas
- Natural Sciences
- Structural Biology
Structural Biology
Structural biology is the branch of biology that asks what biological molecules look like in three dimensions and how those shapes explain what the molecules do.
12 frameworks, from 1950 to today · Biology · Natural Sciences
- Fiber diffraction 1950-1980
- X-ray crystallography 1955-Present
- Model building and refinement 1960-Present
- Structure-function mechanism analysis 1960-Present
- Structural bioinformatics and data archiving 1971-Present
- Molecular dynamics 1977-Present
- Nuclear magnetic resonance spectroscopy of proteins 1985-Present
- Protein structure classification 1990-Present
- Structure-based drug design 1990-Present
- Protein structure prediction 1994-Present
- Integrative structural biology 2005-Present
- Cryo-electron microscopy 2013-Present
Overview
Structural biology is the branch of biology that asks what biological molecules look like in three dimensions and how those shapes explain what the molecules do12. Its central conviction is that function follows form: a protein that splits sugar does so because its surface has a pocket shaped to hold sugar in exactly the right orientation; an antibody neutralises a virus because its binding site matches a patch on the virus's coat1. Where biochemistry studies what molecules react with, structural biology studies the geometry that makes those reactions possible. The field grew out of the meeting of physics, chemistry, and biology in the early twentieth century, and its history is largely a history of instruments and methods—each new way of seeing atoms has opened a new class of questions.
Crystals and the First Diffraction Patterns
The founding technique was X-ray crystallography1. When a beam of X-rays passes through a crystal, the regularly spaced atoms scatter the beam, and the scattered waves interfere to produce a pattern of spots on a detector3. From that pattern, a trained analyst can work back to the arrangement of atoms in the crystal3. The method was developed for simple salts and minerals in the second decade of the twentieth century—Max von Laue showed in 1912 that crystals diffract X-rays, and William Henry Bragg and William Lawrence Bragg worked out the mathematics of interpreting the patterns. The Braggs' law, relating the angle of a diffracted beam to the spacing between planes of atoms, remains the foundation of the technique.
Applying it to biological molecules was far harder, because protein crystals are mostly water and their diffraction is weak. The first protein structure solved by X-ray crystallography was myoglobin, the oxygen-storing muscle protein, by John Kendrew and colleagues in the late 1950s, followed closely by hemoglobin by Max Perutz. These were heroic efforts requiring years of computation with mechanical calculators. The result was a revelation: the protein was not a random tangle but a compact, folded chain with a pocket holding a heme group. The idea that a protein has a definite, reproducible three-dimensional structure—and that this structure is the key to its function—became the organising principle of the field.
The Methods Multiply
Through the 1960s and 1970s, X-ray crystallography remained the dominant method, and the number of solved structures grew slowly. Each structure required growing a crystal, collecting diffraction data, and solving the "phase problem"—the fact that the diffraction pattern records the intensity of scattered waves but not their phase, so the analyst needs extra information to reconstruct the image. Techniques such as multiple isomorphous replacement (soaking heavy atoms into the crystal to change the pattern in a known way) and later molecular replacement (using a known similar structure as a starting model) made the problem tractable.
In parallel, other methods emerged. Fiber diffraction—X-ray diffraction from oriented fibres rather than single crystals—was used by Rosalind Franklin and Maurice Wilkins in the early 1950s to reveal the helical structure of DNA, and later to study fibrous proteins such as collagen and muscle proteins. It gives lower resolution than single-crystal work but can handle molecules that refuse to form proper crystals.
Solution NMR spectroscopy (nuclear magnetic resonance) arrived as a structural method in the 1980s4. Instead of crystallising a molecule, NMR places it in solution and measures the magnetic environment of atomic nuclei, particularly hydrogen. Distances between nearby nuclei can be inferred from how their signals interact, and from those distances a structure can be calculated5. The first protein structure by NMR, of a small protein called proteinase inhibitor IIA, was published by Kurt Wüthrich and colleagues in the mid-1980s. NMR's great advantage is that it studies molecules in solution, closer to their natural state, and can reveal motion; its limitation is size—it works best for smaller proteins.
Cryo-electron microscopy (cryo-EM) is the newest of the three major methods. It images molecules frozen in a thin layer of vitreous ice, using an electron microscope, and averages thousands of individual particle images to build a three-dimensional map. Early cryo-EM in the 1980s and 1990s could resolve only large assemblies at low resolution. The "resolution revolution" came around 2013–2015 with better detectors and improved image-processing software, and cryo-EM began producing atomic-resolution structures of large, flexible complexes that resisted crystallography. It has since become a routine method and has been especially powerful for membrane proteins and large molecular machines.
From Structure to Mechanism
Solving a structure is not the end of the story; the point is to understand how the molecule works. Structure-function mechanism analysis is the interpretive work of relating a shape to a chemical or biological activity: identifying the active site, tracing the path of a substrate, seeing how a conformational change transmits a signal. The classic example is the discovery of the mechanism of serine proteases such as chymotrypsin, where the structure revealed a catalytic triad of three amino acids positioned to split a peptide bond. Similar analyses explained how hemoglobin cooperatively binds oxygen, how antibodies recognise antigens, and how ion channels select for particular ions.
A related development was model building: the painstaking construction of an atomic model that fits the experimental data. In the early decades this was done with wire models and plastic components; today it is done in software, but the principle is the same—the analyst proposes a structure and tests it against the data, adjusting until the model and the evidence agree1. Model building is where the craft of structural biology lives, and errors in it have occasionally led to retractions and corrections.
The Data Explosion and Its Organisation
As the number of structures grew, the field needed ways to store, compare, and classify them. The Protein Data Bank (PDB), established in 1971 at Brookhaven National Laboratory and later managed by an international consortium, became the central archive for experimentally determined structures. Structural bioinformatics and data archiving grew around it: databases, validation tools, and standards for depositing and checking structures. The PDB now holds hundreds of thousands of entries and is a foundation for almost all work in the field.
Protein structure classification schemes such as SCOP (Structural Classification of Proteins) and CATH organise known structures into families, folds, and superfamilies, revealing evolutionary relationships that sequence alone does not show6. Two proteins with no detectable sequence similarity may share a fold, suggesting a common ancestor or a convergently evolved solution to the same problem6.
Computation Enters
Computers changed both how structures are solved and what can be asked of them. Molecular dynamics simulation takes a solved structure and calculates the forces on every atom, then steps the system forward in time to watch it move7. The method was pioneered in the 1970s by Martin Karplus, Michael Levitt, and Arieh Warshel (who shared the 2013 Nobel Prize in Chemistry) and has since become a standard tool for studying protein folding, drug binding, and conformational change. Simulations can now reach biologically relevant timescales for smaller systems, though they remain computationally expensive.
Computational structure prediction asks whether a structure can be predicted from sequence alone, without experiment8. Early efforts in the 1990s and 2000s made progress on small proteins but struggled with the folding problem. The field was transformed by deep-learning methods, above all AlphaFold, developed by DeepMind and announced in 2020, which achieved accuracy competitive with experiment for many proteins. AlphaFold and its successors have made structure prediction a routine first step in biological research, though experimental validation remains essential.
Designing Drugs and Assembling the Whole Picture
Structure-based drug design uses a solved structure to design molecules that fit a target site—for example, an enzyme's active site or a receptor's binding pocket9. The approach produced landmark drugs such as the HIV protease inhibitors in the 1990s, designed with the help of protease structures, and it remains a core method in pharmaceutical research.
Finally, integrative structural biology recognises that many biological systems are too large, flexible, or heterogeneous for any single method. It combines data from crystallography, cryo-EM, NMR, cross-linking mass spectrometry, small-angle X-ray scattering, and computational modelling into a single coherent model, with each method contributing what it does best. This approach has become essential for studying large assemblies such as the nuclear pore complex and the ribosome.
The Present
Structural biology today is a field of multiple complementary methods. Cryo-EM has become the workhorse for large complexes; X-ray crystallography remains indispensable for high-resolution work and for small molecules; NMR continues to provide unique information on dynamics and on proteins in solution; and computational prediction has become a standard companion to experiment. The Protein Data Bank and the classification databases are shared infrastructure. The open questions are substantial: how proteins fold inside cells, how large machines assemble and move, how to model flexibility and disorder, and how to turn structural knowledge into reliable medicines. The field's founding conviction—that shape explains function—has never been more productive.
References
- Structural biology (introduction, History), Wikipedia
- Protein structure, Wikipedia
- X-ray crystallography, Wikipedia
- Triple-resonance nuclear magnetic resonance spectroscopy (Background), Wikipedia
- Nuclear magnetic resonance spectroscopy of proteins, Wikipedia
- Structural Classification of Proteins database, Wikipedia
- Molecular dynamics, Wikipedia
- Protein structure prediction, Wikipedia
- Drug design (Structure-based), Wikipedia
The frameworks of Structural Biology
Each framework is a way of working the field has used: when it held, and the question it answers.
- 1950-1980
- 1955-Present
- 1960-Present
Model building and refinement
How do we build an atomic model that fits the data?
- 1960-Present
Structure-function mechanism analysis
How does molecular shape produce biological function?
- 1971-Present
Structural bioinformatics and data archiving
How should structural data be stored and shared?
- 1977-Present
- 1985-Present
Nuclear magnetic resonance spectroscopy of proteins
What is the structure of a protein in solution?
- 1990-Present
Protein structure classification
How are protein structures related to one another?
- 1990-Present
Structure-based drug design
How can a molecular shape guide drug design?
- 1994-Present
Protein structure prediction
Can a protein's shape be predicted from its sequence?
- 2005-Present
Integrative structural biology
How can many kinds of data be combined into one model?
- 2013-Present
Cryo-electron microscopy
How can we image individual molecules at atomic resolution?
How the ideas connect
Each arrow runs from an earlier idea to the later one it shaped. The list below says what passed between them.
Structure-based drug design grew out of X-ray crystallography
Structure-based drug design grew out of crystallographic structures of drug targets, using the solved geometry of active sites and binding pockets to design inhibitor molecules. HIV protease inhibitors were designed from protease crystal structures in the 1990s.
Carried by HIV protease structures.
Protein structure classification grew out of Structural bioinformatics and data archiving
Protein structure classification schemes such as SCOP and CATH were built directly on the Protein Data Bank archive, using deposited coordinates to group structures into folds and families. The databases made large-scale comparison possible.
Carried by SCOP; CATH; Protein Data Bank (1971).
Structural bioinformatics and data archiving grew out of X-ray crystallography
The Protein Data Bank was established to archive the growing output of X-ray crystallographic structures, making data archiving and structural bioinformatics a direct outgrowth of crystallography. Deposit and validation standards were built around crystallographic data.
Carried by Protein Data Bank (1971); Brookhaven National Laboratory.
Molecular dynamics influenced by X-ray crystallography
Molecular dynamics simulations took crystallographically solved structures as their starting coordinates and used them to study motion and mechanism. The first protein simulation by Karplus and colleagues used a crystallographic structure of a small protein.
Carried by Martin Karplus; Michael Levitt; Arieh Warshel.
Nuclear magnetic resonance spectroscopy of proteins influenced by X-ray crystallography
NMR structure determination adopted concepts and validation criteria from crystallography, and compared its solution structures against crystal structures of the same proteins. The first NMR protein structure was interpreted in light of crystallographic precedent.
Carried by Kurt Wüthrich.
Protein structure prediction influenced by X-ray crystallography
Structure prediction took crystallographically determined structures as its ground truth and its templates, with early methods relying on known folds from crystallography. The goal of prediction was to match experimental crystal structures.
Integrative structural biology influenced by X-ray crystallography
Integrative structural biology uses crystallographic structures as components and restraints in composite models of large assemblies that cannot be crystallised whole. Crystallography supplied both the method of model fitting and many of the input structures.
Cryo-electron microscopy influenced by X-ray crystallography
Cryo-electron microscopy adopted model-building and refinement practices from X-ray crystallography, and its resolution revolution allowed atomic models to be built and validated by the same standards. The two methods became complementary for large complexes.
Cryo-electron microscopy influenced by Model building and refinement
Cryo-EM structure determination relies on model building and refinement software and conventions developed for crystallography, fitting atomic models into density maps. The craft of building and validating models passed directly into cryo-EM practice.
Protein structure prediction influenced by Structural bioinformatics and data archiving
Protein structure prediction methods were trained and benchmarked on structures deposited in the Protein Data Bank, and the PDB archive provided the reference set for evaluating predictions. The data infrastructure of structural bioinformatics made computational prediction possible.
Carried by Protein Data Bank (1971); CASP.
Integrative structural biology influenced by Nuclear magnetic resonance spectroscopy of proteins
Integrative structural biology combines NMR data with crystallography, cryo-EM, cross-linking and scattering, using NMR-derived distance restraints as one input to composite models. The practice of merging restraints from multiple methods grew from the coexistence of NMR and crystallography.
Cryo-electron microscopy rival of Nuclear magnetic resonance spectroscopy of proteins
Cryo-EM and solution NMR competed as methods for studying proteins that resist crystallisation, especially large complexes and membrane proteins. Cryo-EM's resolution revolution made it the dominant alternative for many systems previously tackled by NMR.
Beyond the field
Where Structural Biology's ideas came from in other fields, and what they led to there: the strongest 12 of 48 links.
Fiber diffraction grew out of X-ray crystallography in Materials Science
Fiber diffraction is a branch of X-ray crystallography adapted to partially ordered fibrous materials. Its practitioners used crystallographic theory, Patterson and Fourier methods, and the same instrumentation developed for crystal work.
Carried by William Astbury; Rosalind Franklin.
Fiber diffraction grew out of X-ray diffraction in Materials Characterization
Fiber diffraction applies X-ray diffraction methods to oriented polymer and protein fibers. The technique grew directly from the Laue and Bragg X-ray diffraction work of 1912, using the same diffraction physics and camera methods on fibrous specimens.
Carried by Max von Laue; William Henry Bragg; William Lawrence Bragg.
X-ray crystallography grew out of X-ray diffraction in Materials Characterization
X-ray crystallography grew directly out of X-ray diffraction, using the diffraction of X-rays by crystals as its central method. Bragg's law and the first crystal structures established crystallography as the systematic application of X-ray diffraction.
Carried by William Henry Bragg; William Lawrence Bragg.
Molecular dynamics influenced by Molecular mechanics in Computational Chemistry
Molecular dynamics uses molecular mechanics force fields to compute the forces that drive the simulation. The force-field formalism of molecular mechanics is the foundation on which molecular dynamics trajectories are computed.
Molecular dynamics influenced by Molecular mechanics in Theoretical Chemistry
Molecular dynamics rests on the molecular mechanics force fields developed from the 1940s onward, which supply the potential energy functions integrated in a simulation. Without molecular mechanics there would be no forces to compute.
Nuclear magnetic resonance spectroscopy of proteins grew out of Nuclear magnetic resonance spectroscopy in Spectroscopy
Protein NMR spectroscopy is a branch of NMR spectroscopy: the same physical technique (chemical shift, spin-spin coupling, relaxation) applied to proteins. The first protein spectra were reported in the late 1950s, and the field grew directly out of the established NMR method.
Carried by Felix Bloch; Edward Mills Purcell.
Nuclear magnetic resonance spectroscopy of proteins influenced by Two-dimensional and multidimensional NMR in Spectroscopy
Two-dimensional NMR, developed by Ernst and colleagues with COSY and NOESY experiments, supplied the sequential-assignment and distance-restraint methods that made protein structure determination by NMR possible. Protein NMR is unthinkable without these pulse sequences.
Carried by Richard R. Ernst; Kurt Wüthrich.
Protein structure classification grew out of Structural Biology in Biochemistry
Protein structure classification (SCOP, CATH) grew directly out of structural biology: it systematised the growing body of solved protein structures into hierarchies of folds and families. It is a branch of structural biology's comparative work.
Carried by Alexey Murzin; Cyrus Chothia.
Structure-based drug design influenced Fragment-based lead discovery in Drug Discovery
Structure-based drug design, especially crystallographic and NMR fragment soaking, provided the structural methods by which fragment hits are detected and grown into leads.
Structure-based drug design influenced Fragment-based lead discovery in Medicinal Chemistry
Structure-based drug design supplied the crystallographic and NMR methods for observing fragment binding in target sites and the strategy of growing fragments into leads.
Structure-based drug design influenced Structure-Based Receptor Pharmacology in Receptor Theory And Signal Transduction
Structure-based drug design supplied the crystallographic and computational methods that structure-based receptor pharmacology applies to receptor-ligand complexes. The later field is the extension of these methods from enzymes and general targets to receptor pharmacology.
Structure-based drug design led to Structure-based drug design and molecular docking in Computational Chemistry
Structure-based drug design and molecular docking is the computational branch that grew out of structure-based design: docking programs score and pose candidate ligands in a known binding site. It is the algorithmic realisation of the structure-based design idea.

Structural Biology is a planet in Nousfarer
Its city has a tower for each of these 12 frameworks. Land, walk in, and climb one concept at a time.
Land on Structural BiologyRelated fields
Fields its ideas are linked to
- Biochemistry 9 links
- Biophysics 8 links
- Cell Biology 4 links
- Computational Chemistry 3 links
- Materials Characterization 3 links
- Drug Discovery 3 links
- Molecular Biology 2 links
- Medicinal Chemistry 2 links
- Spectroscopy 2 links
- Mineralogy And Petrology 2 links
- Chemical Thermodynamics 1 link
- Organic Chemistry 1 link
More in Biology
- Biochemistry
- Biology
- Cell Biology
- Cell Signaling
- Developmental Biology
- Genetics
- Molecular Biology
- Physiology
- Synthetic Biology
- Systems Biology
See where Structural Biology sits on the Map of Human Knowledge