Measure compactness
Rg summarizes how widely selected atoms are distributed around their center of mass.
What Rg measures, how compactness is calculated around the center of mass, how to interpret increasing or decreasing values, when it is useful, and when it can mislead you.
Rg summarizes how widely selected atoms are distributed around their center of mass.
One Rg value per trajectory frame reveals compaction, expansion, or stable spatial size over time.
Similar Rg values can belong to different shapes, so structural visualization and companion analyses remain essential.
Radius of gyration, usually written as Rg, is a geometric measure of how far the mass of a molecular structure is distributed from its center of mass.
Rg answers this question: “How tightly or widely are the selected atoms distributed around their common center?” A compact structure generally has a smaller Rg than an expanded structure made from the same selected atoms.
Rg is normally reported in nanometres (nm) or ångströms (Å). Remember: 1 nm = 10 Å.
Rg compresses the three-dimensional spread of many atoms into one interpretable measure of global compactness.
Shows whether the selected structure remains spatially compact, becomes more compact, or expands.
Helps detect global unfolding, domain separation, polymer collapse, or compaction events.
Supports consistent comparison of apo and bound proteins, mutants, conditions, or replicate simulations.
No. Molecular dynamics generates atomic coordinates over time. Radius of gyration is an analysis metric calculated from a structure or from every saved frame of that trajectory.
Move the compactness slider to see atoms approach or move away from the center of mass, while the enclosing Rg circle and calculated value update live.
A teaching model showing how the same atom group can be compact or expanded.
The selected atoms occupy a moderate spatial region around the center of mass. The Rg value reflects their root-mean-square spread, not the outermost molecular boundary.
See how atom mass and distance from the center contribute to the final Rg value.
Here, dᵢ is the distance from atom i to the center of mass. Atoms with larger masses contribute more when mass weighting is used.
| Atom | Mass | d (Å) | m·d² |
|---|
Correct molecule reconstruction and atom selection are as important as the equation.
Choose the protein, backbone, Cα atoms, domain, complex, polymer, ligand, or another biologically meaningful atom group.
Correct periodic-boundary wrapping so one molecule is not artificially split across opposite sides of the simulation box.
For the selected atoms, calculate the mass-weighted average position. Some software can use other weights, but the method must be reported.
For each atom, calculate its three-dimensional distance from the center of mass in the same frame.
Square each distance, apply the chosen weight, divide by the total weight, and take the square root.
Calculate one Rg value per frame and plot Rg against simulation time.
The selected mass is concentrated closer to its center of mass. For the same system and selection, this usually indicates a more compact spatial distribution.
The selected mass is distributed farther from its center of mass. This may indicate expansion, extended loops, domain separation, unfolding, or a naturally elongated shape.
Measures compactness using all non-hydrogen protein atoms. It captures both backbone and side-chain packing but may be more sensitive to side-chain rearrangements.
Useful for monitoring the global fold while reducing sensitivity to side-chain motion. Define exactly which backbone atoms were used.
Provides a simplified representation with one atom per residue. It is useful for consistent comparisons but is not identical to an all-atom mass distribution.
Useful for multi-domain proteins where global Rg may be dominated by relative domain movement. Domain-wise Rg can reveal local compactness separately.
A combined complex Rg can be calculated, but ligand position and protein size may make interpretation difficult. Protein-only Rg is usually reported separately.
For a flexible ligand or peptide, Rg can describe internal extension or collapse. It does not reveal where the ligand sits in the binding pocket.
Rg is useful for DNA/RNA, polymers, micelles, and assemblies, but the selected group and axis interpretation must match the system geometry.
These patterns are clues, not automatic biological conclusions.
Suggests the selected group maintains a broadly consistent spatial compactness. It does not prove complete equilibrium or unchanged conformation.
May indicate compaction, folding, collapse, domain closure, or tighter packing. Confirm with SASA, contacts, secondary structure, and visualization.
May indicate expansion, domain opening, loss of compact contacts, or partial unfolding. It can also be a legitimate functional transition.
May reflect breathing motions, flexible domains, repeated opening and closing, state exchange, inadequate sampling, or a PBC problem.
Could be a true rapid transition, complex dissociation, or an artefact from a molecule split by periodic boundaries. Inspect the trajectory immediately.
Small differences may be meaningful or may fall within normal variability. Use replicates, block statistics, and uncertainty instead of judging curves visually.
| Observation | Possible interpretation | What must be checked | বাংলায় সংক্ষেপ |
|---|---|---|---|
| Rg decreases | Compaction, folding, collapse, domain closure, stronger packing, or loss of extended tails. | SASA, native contacts, secondary structure, domain distance, and visualization. | Structure compact হতে পারে। |
| Rg increases | Expansion, unfolding, domain opening, loop extension, dissociation, or elongated functional state. | RMSD, secondary structure, contacts, SASA, interdomain distances, and PBC. | Structure প্রসারিত হতে পারে। |
| Rg unchanged | Overall spatial size remains similar. | The conformation may still change internally; inspect RMSD, PCA, contacts, and shape. | Total size একই হলেও shape বদলাতে পারে। |
| Rg fluctuates | Normal breathing, flexible regions, opening/closing, state transitions, or insufficient sampling. | Amplitude, timescale, replicates, trajectory visualization, and axis components. | Breathing motion বা state change হতে পারে। |
Larger proteins usually have larger absolute Rg values, so direct comparison across different sizes can be misleading.
Elongated or multi-domain structures can have high Rg even when well folded and stable.
Termini and disordered regions can strongly increase both mean Rg and fluctuation.
Binding or mutation may close, open, rigidify, or destabilize a protein architecture.
Environmental conditions influence packing, expansion, hydration, and conformational sampling.
Association or dissociation changes the spatial mass distribution of a complex.
Cα, backbone, heavy atoms, domains, and full complexes produce different numerical values.
A molecule split across the box can generate an artificial and extremely large Rg.
Mass, unit, or other weights alter the contribution of individual atoms to the result.
The total Rg gives one overall spread value. It is rotationally invariant, meaning simple rotation of the same conformation does not change the total Rg.
Uses distances in the y–z plane. It describes mass spread perpendicular to the x-axis.
Uses distances in the x–z plane. It describes mass spread perpendicular to the y-axis.
Uses distances in the x–y plane. It describes mass spread perpendicular to the z-axis.
A compact globular structure and a differently arranged structure can share a similar total Rg. Shape descriptors such as the gyration tensor eigenvalues, asphericity, acylindricity, relative shape anisotropy, domain distances, or principal-axis analysis provide additional information.
Rg measures compactness around the center of mass. RMSD measures coordinate deviation from a reference after fitting.
Rg is global. RMSF shows how strongly individual atoms or residues fluctuate around an average or reference position.
Rg measures spatial spread of mass. SASA measures surface accessible to a solvent probe.
Rg uses all selected atoms. End-to-end distance uses only two terminal points and can miss internal compaction.
Rg can change without specifying which contacts form or break. Native-contact analysis provides structural detail.
Rg gives one compactness coordinate. PCA and clustering can identify different conformational states with similar Rg.
This can place parts of one molecule far apart and create a meaningless large Rg. Reconstruct the molecule and visually verify the trajectory.
Cα, backbone, heavy-atom, and whole-complex Rg are not interchangeable. Use the same selection for comparisons.
A lower Rg only indicates greater compactness. A compact misfolded state can have low Rg, and a stable elongated protein can have high Rg.
Absolute Rg values depend strongly on system size and composition. Compare related systems or use suitable normalized/scaling approaches.
Cartesian components can change when the molecule rotates. Use a meaningful frame, principal axes, or an aligned coordinate system.
Long tails can dominate Rg changes even when the structural core is stable. Report both whole-protein and core/domain Rg when appropriate.
A single trajectory may not represent the underlying distribution. Independent replicates improve confidence in observed differences.
Report distributions, averages with uncertainty, block behaviour, and replicate variation instead of judging only line separation.
Decide whether you need whole-protein compactness, core compactness, domain behaviour, peptide collapse, or assembly size.
Use the same composition and selection across compared systems.
Make the molecule whole and confirm that no frame is artificially split.
Start with total Rg. Add domain-wise or axis-aware analysis only when scientifically meaningful.
Evaluate time evolution, histograms, transitions, block averages, and replicate variation.
Use visualization, RMSD, RMSF, SASA, contacts, secondary structure, PCA, clustering, and relevant distances.
The exact PBC workflow depends on whether the system is soluble, membrane-bound, oligomeric, or contains multiple molecular groups.
gmx trjconv -s md.tpr -f md.xtc -o md_nojump.xtc -pbc nojump
Select the molecular system or appropriate group. Visually verify that the molecule remains continuous.
gmx trjconv -s md.tpr -f md_nojump.xtc -o md_center.xtc -pbc mol -center -ur compact
Select the protein or complex for centering, then the system for output. Verify the result carefully, especially for membranes and multimers.
gmx gyrate -s md.tpr -f md_center.xtc -o gyrate.xvg
Choose the desired analysis group, such as Protein, Backbone, C-alpha, or a custom domain group. GROMACS reports total Rg and components about the x, y, and z axes as a function of time.
gmx make_ndx -f md.tpr -o index.ndx gmx gyrate -s md.tpr -f md_center.xtc -n index.ndx -o gyrate_domain.xvg
Create biologically meaningful groups and document residue ranges clearly.
“The mass-weighted radius of gyration of the selected protein heavy atoms was calculated for each production-trajectory frame after periodic-boundary reconstruction. Total Rg values were reported in nanometres as a function of simulation time and compared across independent replicas.”
It supports the conclusion that global compactness is consistent, but it cannot prove correct folding, thermodynamic stability, convergence, or biological function.
No. A compact misfolded structure may have a low Rg, while a naturally elongated and stable protein may have a larger Rg.
First check whether the molecule is split across periodic boundaries or whether an oligomer dissociated. Then investigate a genuine conformational event.
Total scalar Rg does not require rotational or translational fitting. It does require a correctly reconstructed molecular group. Axis-specific interpretation may require a meaningful orientation.
They use different spatial and mass distributions. Side chains contribute to heavy-atom Rg but are absent from a Cα-only calculation.
Yes. Rg reduces the structure to one spread value. Different shapes and contact patterns can share the same Rg.
Yes, especially for flexible ligands or peptides, but it describes internal compactness rather than binding-site location or binding strength.
Usually not when studying protein compactness, because mobile solvent and ions would dominate or distort the selected mass distribution.
Radius of gyration tells you how widely the selected molecular mass is distributed around its center of mass. It does not independently prove structural stability, correct folding, ligand binding, convergence, or biological activity.
Educational content for AMRA-LAB Computational Analyses · Radius of Gyration module