TrialLineage Concept
Structural biology
Before scientists can design a drug to fit a protein, they need to see the protein. Structural biology is the field that makes that possible — determining the three-dimensional shapes of biological molecules at atomic or near-atomic resolution. This page explains what structural biology is, why it matters to drug discovery, and how it became essential to the development of KRAS-directed therapies now entering clinical trials in pancreatic cancer.
In plain language
What is structural biology?
Structural biology is the science of figuring out the physical shape of proteins, nucleic acids, and other large molecules inside living systems. Proteins are not flat. They fold into complex three-dimensional forms — with grooves, pockets, hinges, and surfaces — and those shapes determine what the protein can do. A protein’s function depends on its structure: which molecules it can bind, how it changes shape when activated, and where a drug might be able to interfere.
Structural biologists use specialized techniques to produce detailed maps of these shapes, typically at resolutions measured in angstroms (tenths of a nanometer). These maps are not photographs — they are computational reconstructions built from physical measurements. The result is a model that shows where every atom in a protein sits, how the protein moves, and what openings exist on its surface. For drug discovery, this information is foundational: it turns a protein from an abstract target into a physical object that chemists can design molecules to fit.
Key methods
- X-ray crystallography: the protein is crystallized and X-rays are passed through it; the diffraction pattern reveals its atomic structure — for decades the dominant method in the field
- Cryo-electron microscopy (cryo-EM): the protein is flash-frozen and imaged by an electron beam; computational averaging of many images produces a high-resolution structure without needing crystals
- NMR spectroscopy: uses magnetic fields to study the structure and dynamics of proteins in solution, particularly useful for small or flexible proteins
- Computational structure prediction: algorithms such as AlphaFold predict protein structures from amino acid sequences, increasingly complementing experimental methods
Why it matters
Why structural biology matters in biomedical discovery
It turns targets from names into objects
When geneticists identify a gene linked to disease, and cell biologists show that its protein product drives a pathological process, the target is still an abstraction for drug designers. Structural biology provides the physical map: where the protein’s active sites are, how it interacts with other molecules, and what surfaces might be exploitable by a drug. Without this information, medicinal chemistry is largely guesswork.
It explains why some targets are hard to drug
Structural data often reveals why a target resists chemical intervention. A protein with a smooth, featureless surface offers no pocket for a small molecule to bind. A protein that binds its natural partner with extreme affinity leaves little room for a competitor. These are structural problems, visible only when the protein’s shape is known — and they define the strategic challenges that medicinal chemistry must solve.
It enables rational drug design
When chemists can see the exact shape of a binding pocket on a protein, they can design molecules to fit it — adjusting the molecule’s shape, charge, and flexibility to complement the pocket’s geometry. This is called structure-based or rational drug design, and it is substantially more efficient than testing compounds at random. Many of the most important drugs developed over the past three decades owe their existence to structural data that guided their design.
Understanding function through form
How structural biology reveals what proteins are actually doing
Proteins are not static. They switch between conformations, open and close binding sites, and change shape in response to signals. Structural biology captures these states and helps scientists understand how a protein’s behavior connects to disease.
Active vs. inactive states
Many disease-driving proteins, including KRAS, cycle between active and inactive conformations. Structural biology can resolve both states, revealing differences in surface shape and pocket availability. Understanding which state a drug needs to engage — and when that state is accessible — is critical to designing effective inhibitors.
Mutation effects
When a gene mutation changes a single amino acid in a protein, the structural consequences can be subtle or dramatic. Structural biology shows exactly how a mutation alters the protein’s shape, its interactions with other molecules, and whether it creates or destroys potential drug-binding sites. For KRAS, different mutations (G12C, G12D, G12V) produce slightly different structural landscapes, which is why drugs designed for one variant may not work on others.
Protein-protein interactions
Proteins rarely act alone. KRAS, for example, transmits signals by physically binding to downstream effectors like RAF. Structural biology can map the interface where two proteins meet, showing the contact surfaces, buried residues, and conformational changes that occur upon binding. This information is essential for understanding signaling and for evaluating whether a protein-protein interface could be disrupted by a drug.
Connection to KRAS, daraxonrasib, and pancreatic cancer
Why structural biology was decisive for KRAS drug development
The switch-II pocket changed everything
For decades, KRAS was considered structurally undruggable. Its surface appeared smooth, with no deep pocket suitable for a small molecule. The pivotal advance came when structural studies — including work from the Shokat lab at UCSF — revealed a previously unrecognized pocket near the switch-II region of KRAS G12C. This pocket is not visible in the protein’s active state; it only opens in the inactive, GDP-bound conformation. Without structural biology, this vulnerability would not have been found.
Structure guided the design of covalent inhibitors
Once the switch-II pocket was identified, medicinal chemists used structural data to design molecules that could fit into it and form a covalent bond with a specific cysteine residue (Cys12 in KRAS G12C). The shape, orientation, and chemical environment of the pocket — all determined by structural biology — dictated what the drug molecule had to look like. Each round of chemical optimization was guided by new co-crystal structures showing how the drug and protein interact at atomic resolution.
Extending to other KRAS mutations requires new structural work
The covalent strategy that works for KRAS G12C depends on the presence of a reactive cysteine at position 12 — a feature unique to that mutation. KRAS G12D and G12V, which are more prevalent in pancreatic cancer, have different amino acids at that position and different local geometries. Structural biology is now being used to identify alternative binding sites and inhibitor strategies for these variants. The structural landscape of each mutant form is distinct, and new drug designs must be guided by new structural data.
Daraxonrasib’s binding mode is a structural story
The specific way daraxonrasib engages its target — which pocket it occupies, how it orients within that pocket, what contacts it makes with surrounding residues — is defined by structural biology. Co-crystal structures of the drug bound to its target are what allow scientists to understand why it is selective for one mutant form, how resistance might develop, and how future compounds could be improved. Structure is not merely background context for daraxonrasib; it is the basis of the drug’s design.
Branch points in scientific thinking
How structural biology thinking branched and evolved
The field has not followed a single methodological path. Several branching decisions in how structural biology is practiced have directly influenced the KRAS drug discovery story.
Crystallography vs. cryo-EM
Grow crystals, or freeze individual molecules?
X-ray crystallography dominated structural biology for decades and produced the foundational structures of KRAS. But crystallography requires the protein to form ordered crystals — a process that can be difficult, slow, or impossible for some conformational states. Cryo-EM, which advanced dramatically after the mid-2010s, can determine structures without crystals, opening up access to flexible proteins, transient conformations, and large complexes that crystallography struggles with. Both methods continue to be used, often complementarily.
Static snapshots vs. conformational dynamics
One structure, or many?
Early structural biology tended to treat a protein’s structure as a single, fixed form. But proteins are dynamic: they breathe, flex, and shift between conformational states. A growing branch of the field focuses on capturing multiple conformations and understanding how the protein moves. For KRAS, this shift was important — the switch-II pocket is only present in certain conformational states, and a static view of the protein would not have revealed it. Techniques like time-resolved crystallography, molecular dynamics simulations, and cryo-EM classification of multiple states are now standard tools.
Experimental structures vs. computational prediction
Measure in the lab, or predict from sequence?
The emergence of highly accurate computational structure prediction — most prominently through AlphaFold — has created a new branch in how structural information is obtained. For well-studied proteins like KRAS, experimental structures remain the gold standard for drug design because they capture the fine detail of drug-binding pockets and conformational states. But computational methods are increasingly useful for related proteins, for generating hypotheses about mutant-specific structures, and for targets where experimental structures are not yet available.
Limitations and incomplete approaches
Structural approaches that fell short — but still advanced understanding
Structural biology has not always delivered the answers that drug discovery needed, when it needed them. Several limitations and partial successes shaped the KRAS story.
Early KRAS structures showed no druggable pocket
The first crystal structures of KRAS, solved in the 1980s and 1990s, showed a relatively small, smooth protein with a tightly bound nucleotide and no obvious pocket for a drug. These structures were scientifically valuable — they revealed the protein’s fold, GTPase mechanism, and interaction surfaces — but they reinforced the view that KRAS was structurally intractable. It took decades and new approaches to reveal the hidden pockets that eventually enabled drug design.
Crystal packing artifacts obscured flexible regions
In crystallography, the process of crystallization can lock a protein into one conformation and obscure regions that are flexible in solution. For KRAS, some of the most therapeutically relevant features — particularly the switch-II loop — are intrinsically flexible and can be poorly resolved or constrained in crystal structures. This meant that the binding pocket exploited by covalent inhibitors was not apparent in early crystal forms, even though it exists in the protein’s natural conformational landscape.
Structures of GTP-bound (active) KRAS were less useful for drug design
Much early structural work focused on the active, GTP-bound form of KRAS — the state that drives signaling. But the druggable switch-II pocket is accessible only in the inactive, GDP-bound state. This mismatch between the biologically interesting state and the chemically exploitable state delayed progress. The field eventually recognized that solving structures of multiple nucleotide-bound and mutant-specific states was necessary to find druggable opportunities.
Resolution limitations in early cryo-EM
Before the hardware and software advances of the mid-2010s (sometimes called the “resolution revolution”), cryo-EM could not reliably resolve proteins as small as KRAS at the atomic detail needed for drug design. This meant that crystallography remained the only viable route for KRAS structures for many years. As cryo-EM resolution has improved, it has become increasingly applicable to small GTPases and their complexes, expanding the structural toolkit for KRAS-related programs.
What often gets missed
What the public usually does not hear about structural biology
Structural biology is rarely mentioned in public accounts of drug discovery. When a new drug is announced, the structural work that made it possible is almost never explained — even though it is often the single most important enabling step.
Structures are not photographs
The colorful images of proteins that appear in scientific papers and news stories are not photographs. They are computational models built from diffraction patterns, electron micrographs, or NMR spectra. Producing them requires months or years of sample preparation, data collection, and computational refinement. The impression of simplicity that these images convey conceals enormous technical effort.
Solving a structure does not mean the problem is solved
Knowing a protein’s shape is necessary but not sufficient for designing a drug against it. The structure shows where a pocket is, but not whether a drug-like molecule can be built to fill it. Many proteins have been structurally characterized in great detail and remain undrugged — not because the structure is unknown, but because the chemistry to exploit it has not been achieved.
Drug design is an ongoing conversation with structure
In modern drug development, structural biology is not a one-time input. Medicinal chemists and structural biologists work in iterative cycles: chemists design a compound, structural biologists solve the structure of the compound bound to the protein, and the team uses the result to guide the next round of chemical design. This cycle — sometimes repeated dozens of times — is how drugs like KRAS inhibitors are optimized.
The field has changed rapidly in the past decade
Advances in cryo-EM hardware, computational structure prediction, and high-throughput crystallography have transformed what structural biology can deliver. Structures that once took years can now sometimes be obtained in weeks. This acceleration has direct implications for drug discovery — it means that new targets can be structurally characterized faster, and that drug design cycles can move more quickly from hypothesis to atomic-resolution feedback.
Related case
Where this concept appears in TrialLineage
Daraxonrasib in pancreatic cancer
Structural biology sits at a critical junction in the scientific lineage behind daraxonrasib. It provided the three-dimensional insight that turned KRAS from a biologically validated but chemically intractable target into one that medicinal chemists could design drugs against. The case page traces the full discovery chain — from oncogene discovery through signaling biology, disease research, structural insight, medicinal chemistry, and clinical translation — showing how a phase 1–3 trial emerges from decades of interrelated work.
Related concepts
Other scientific fields in the TrialLineage discovery chain.
About this page
This is a TrialLineage concept explainer. Concept pages provide plain-language background on the scientific fields, branch points, and discoveries that underlie specific clinical developments. They are designed to be read independently or as companions to case pages — helping a public audience understand the full discovery process behind a human-disease trial.