# How to Analyze Molecular Dynamics Trajectories: A Practical Guide to RMSD, RMSF, and Radius of Gyration

Molecular dynamics simulations generate trajectories that depict a system's evolution over time, and researchers analyze these trajectories both visually and quantitatively to characterize protein behavior. The most commonly conducted analyses include root mean square deviation (RMSD), radius of gyration (Rg), and root mean square fluctuation (RMSF), though these methods are limited by their strictly statistical nature. This guide explains what each metric measures, how to compute it from your simulation output, and how to interpret the results in the context of protein stability, flexibility, and compactness. You will learn which metrics to prioritize for different research questions, how to prepare your trajectory files for analysis, and how to recognize when your simulation has not converged or when your analysis protocol needs adjustment.

## What Molecular Dynamics Trajectories Contain and Why Analysis Matters

A molecular dynamics simulation produces a trajectory file that records the positions of every atom in your system at regular time intervals. For a typical protein simulation, this means thousands of frames, each containing the three-dimensional coordinates of every atom in the protein, solvent, ions, and any ligands present. The trajectory is the raw material for all downstream analysis, and the quality of your conclusions depends directly on how carefully you process and interpret this data.

The trajectory captures the system's evolution in time, and the analysis of that trajectory is what turns raw coordinates into biological insight. Without quantitative analysis, a trajectory is simply a large collection of coordinates that is difficult to interpret by eye. Metrics like RMSD, RMSF, and Rg reduce this complex data into interpretable numbers that tell you whether your protein remained stable, which regions moved most, and whether the overall structure stayed compact.

For researchers working in structural bioinformatics, protein structure prediction, and molecular docking interpretation, trajectory analysis is the step that validates whether your computational model behaves plausibly. A docking study that identifies a promising binding pose is only meaningful if the protein-ligand complex remains stable over time, and that stability is assessed through trajectory metrics. Similarly, a homology model is only credible if it maintains its fold during simulation, which RMSD and Rg measurements can confirm.

## At a Glance: Choosing the Right Trajectory Metric

The table below summarizes the three core trajectory metrics, what each one measures, and the practical questions each metric answers. Use this table to decide which analysis to run first based on your research question.

| Metric | What It Measures | Primary Question Answered | Typical Use Case |
|--------|------------------|---------------------------|------------------|
| RMSD | Deviation of atomic positions from a reference structure over time | Did the protein remain close to its starting structure? | Assessing overall stability and convergence of the simulation |
| RMSF | Per-residue fluctuation around the average structure | Which residues or regions move the most? | Identifying flexible loops, terminal regions, and binding site dynamics |
| Radius of Gyration | Distribution of atoms around the center of mass | Did the protein stay compact or expand over time? | Evaluating global folding state and structural compactness |

These three metrics are complementary. RMSD gives you a global view of structural stability, RMSF gives you a local view of residue-level flexibility, and Rg gives you a view of overall compactness. A complete trajectory analysis typically includes all three, and the combination of their results provides a more complete picture than any single metric alone.

## Core Principles of Trajectory Analysis

### Reference Structure Selection and Its Impact on Results

The choice of reference structure is the first decision that affects all subsequent RMSD calculations. The reference structure is the frame against which all other frames are compared, and different reference choices produce different RMSD values. The most common choices are the first frame of the simulation, the energy-minimized starting structure, or the average structure calculated over the entire trajectory.

Using the first frame as a reference shows how far the protein has moved from its initial conformation. This is useful for assessing whether the simulation has equilibrated, because a stable simulation should show RMSD values that plateau after an initial rise. Using the average structure as a reference removes the bias of the starting conformation and shows fluctuations around the central structure, which is more appropriate for assessing equilibrium dynamics.

For protein-ligand complexes, you must decide whether to calculate RMSD on the protein alone, the ligand alone, or the entire complex. Each choice answers a different question. Protein RMSD tells you whether the receptor remained stable, ligand RMSD tells you whether the ligand stayed in its binding pose, and complex RMSD gives an overall view. Most published studies report protein backbone RMSD and ligand RMSD separately, because combining them can obscure important details.

### Superposition and Alignment Before RMSD Calculation

RMSD calculation requires that the structures being compared are properly aligned. Without superposition, the RMSD value includes both rotational and translational differences between frames, which inflates the value and makes it meaningless for assessing internal structural changes. The standard practice is to perform a least-squares superposition on a stable subset of atoms, typically the protein backbone atoms or C-alpha atoms, before calculating RMSD.

The choice of atoms used for superposition matters. If you superimpose on all atoms, including flexible side chains and loops, the alignment will be dominated by the most mobile regions and may not reflect the overall structural relationship. Superimposing on backbone atoms or C-alpha atoms is more robust because these atoms are less mobile than side chains. For multi-domain proteins, you may need to superimpose on individual domains separately to assess domain movements.

Most trajectory analysis tools, including GROMACS, AmberTools, and MDAnalysis, perform superposition automatically when you request RMSD calculations. However, you should verify that the default settings match your analysis needs. Some tools default to fitting on all atoms, which may not be appropriate for your system.

### Trajectory Trimming and Equilibration Assessment

The first portion of a simulation trajectory typically represents the equilibration phase, during which the system adjusts from the initial conditions to a stable state. Including this phase in your analysis can skew your results, because the early frames may show large structural changes that do not reflect equilibrium behavior.

The standard approach is to discard the equilibration period before analysis. The length of this period depends on the system size, the starting structure quality, and the simulation parameters. For a well-prepared system, equilibration may take only a few nanoseconds, while for a poorly prepared system or a large complex, it may take tens of nanoseconds. You can assess whether equilibration is complete by plotting RMSD against time and looking for a plateau. Once the RMSD values stabilize, the system is considered equilibrated.

Some researchers prefer to run a separate equilibration phase and then begin the production run from the equilibrated state. This approach separates the equilibration dynamics from the production dynamics and makes the analysis cleaner. If you use this approach, you should still verify that the production run is stable by checking RMSD values over the production trajectory.

## Practical Workflow for Computing RMSD, RMSF, and Rg

### Step 1: Prepare Your Trajectory Files

Before you can compute any metric, you must ensure that your trajectory files are complete and properly formatted. Most simulation packages produce trajectory files in their own format, such as GROMACS XTC or TRR files, Amber NetCDF files, or Desmond CMS files. Your analysis tool must be able to read both the trajectory file and the corresponding topology or structure file.

The topology file is essential because it defines the atoms, residues, and bonds in your system. Without the topology, the analysis tool cannot interpret the coordinates in the trajectory file. Verify that the topology and trajectory files correspond to the same system and that the atom ordering matches. A common error is using a trajectory from one simulation with a topology from another, which produces meaningless results.

If your trajectory is very large, you may need to reduce the number of frames for analysis. Most analysis tools allow you to skip frames at regular intervals, such as analyzing every 10th or 100th frame. This reduces the computational cost of analysis without significantly affecting the results, provided that the skipped frames do not contain important events. For most analyses, analyzing frames at 10 to 100 picosecond intervals is sufficient.

### Step 2: Select the Appropriate Analysis Tool

Several software packages can compute RMSD, RMSF, and Rg from molecular dynamics trajectories. The choice of tool depends on your simulation package, your familiarity with the software, and the complexity of your analysis.

GROMACS provides the gmx rms, gmx rmsf, and gmx gyrate commands for these analyses. These commands are well documented and widely used, and they integrate directly with GROMACS trajectory and topology files. The GROMACS tools are command-line based, which makes them suitable for scripting and reproducible analysis workflows.

MDAnalysis and MDTraj are Python libraries that provide flexible trajectory analysis capabilities. These libraries allow you to write custom analysis scripts and integrate trajectory analysis with other Python-based data processing. They are particularly useful if you need to perform custom analyses beyond the standard metrics or if you want to generate publication-quality plots with Matplotlib.

AmberTools provides the cpptraj program, which can compute RMSD, RMSF, and Rg from Amber trajectories. cpptraj is highly flexible and can handle a wide range of trajectory formats. It is the standard analysis tool for Amber simulations.

For researchers who prefer graphical interfaces, VMD provides trajectory analysis capabilities through its graphical interface and scripting language. VMD is particularly useful for visualizing trajectories and for interactive exploration of simulation results.

### Step 3: Compute RMSD for Overall Stability Assessment

To compute RMSD, you need to specify the reference structure, the atoms to include in the calculation, and the atoms to use for superposition. The standard choice is to calculate backbone RMSD using the first frame or the energy-minimized structure as the reference.

The RMSD calculation produces a value for each frame in the trajectory, which you can plot against time. A stable simulation shows RMSD values that rise initially and then plateau at a consistent level. The plateau value depends on the protein size and flexibility, but for a globular protein, backbone RMSD values of 1 to 3 angstroms are typical for stable simulations. Values above 5 angstroms may indicate significant conformational changes or instability.

For protein-ligand complexes, you should calculate ligand RMSD separately. Ligand RMSD is calculated after superimposing the protein onto the reference structure, which removes the protein's overall motion and reveals the ligand's movement relative to the binding site. A stably bound ligand shows low RMSD values, typically below 2 angstroms, throughout the simulation. Large ligand RMSD values indicate that the ligand has moved from its initial binding pose.

### Step 4: Compute RMSF for Residue-Level Flexibility

RMSF is calculated per residue and shows the average fluctuation of each residue around its average position. The calculation requires that you first superimpose all frames onto a reference structure, typically the average structure, to remove overall translational and rotational motion. Then, for each residue, you calculate the root mean square fluctuation of its atoms across all frames.

The RMSF calculation produces a value for each residue, which you can plot against residue number. High RMSF values indicate flexible regions, while low values indicate rigid regions. In a typical protein, the N-terminal and C-terminal regions show high RMSF values because they are less constrained by tertiary interactions. Loops and turns also show higher RMSF values than secondary structure elements. Alpha helices and beta sheets generally show low RMSF values because they are stabilized by hydrogen bonding networks.

RMSF analysis is particularly useful for identifying flexible regions that may be involved in function. For example, a binding site located in a flexible loop may show higher RMSF values than the rest of the protein, indicating that the loop undergoes conformational changes upon ligand binding. Comparing RMSF values between apo and holo simulations can reveal ligand-induced changes in flexibility.

### Step 5: Compute Radius of Gyration for Compactness Assessment

The radius of gyration measures the distribution of atoms around the center of mass. A compact protein has a low Rg value, while an extended or unfolded protein has a high Rg value. The Rg is calculated for each frame and plotted against time to show how the protein's compactness changes over the simulation.

For a stable globular protein, Rg values should remain relatively constant throughout the simulation. A gradual increase in Rg over time may indicate that the protein is unfolding or expanding, while a decrease may indicate compaction. The expected Rg value depends on the protein size and shape. For a globular protein, Rg is typically proportional to the cube root of the molecular weight.

Rg analysis is particularly useful for assessing whether a protein maintains its folded state during simulation. If you are studying a protein that is known to be globular, a stable Rg value confirms that the simulation preserves the folded state. If you are studying protein folding or unfolding, Rg changes over time provide direct evidence of conformational transitions.

## Options and Tradeoffs in Trajectory Analysis

### Choosing Between Backbone and All-Atom Analysis

The choice between backbone and all-atom analysis affects the sensitivity of your metrics. Backbone RMSD and RMSF are less sensitive to side chain movements and provide a cleaner view of the protein's overall fold. All-atom analysis includes side chain movements, which can be substantial for surface residues and may obscure the overall structural picture.

For most stability assessments, backbone analysis is preferred because it focuses on the protein's secondary and tertiary structure. Side chain movements are often not relevant to questions of overall stability. However, if you are studying side chain conformations, such as in a binding site, all-atom analysis may be more appropriate.

The choice also affects the computational cost. All-atom analysis requires more memory and processing time than backbone analysis, particularly for large proteins. For a typical analysis, the difference is small, but for very large systems or very long trajectories, it may become significant.

### Comparing Multiple Simulations and Replicates

Many research questions require comparing multiple simulations, such as comparing an apo protein with a protein-ligand complex or comparing multiple ligand candidates. When comparing simulations, you must ensure that the analysis parameters are identical across all simulations. This includes the reference structure, the atoms used for superposition, and the frame selection.

For comparing RMSD values between simulations, you should use the same reference structure for all simulations. If you use the first frame of each simulation as the reference, the RMSD values will reflect the deviation from each simulation's starting structure, which may differ between simulations. Using a common reference, such as the crystal structure or the energy-minimized structure, allows direct comparison.

For comparing RMSF values, you should calculate the average structure for each simulation separately and then compare the fluctuation patterns. Differences in RMSF between simulations can reveal ligand-induced changes in flexibility or differences in stability between mutants.

### Handling Multi-Domain Proteins and Large Complexes

Multi-domain proteins present special challenges for trajectory analysis. The overall RMSD of a multi-domain protein may be dominated by inter-domain movements, which can obscure the stability of individual domains. Similarly, the Rg of a multi-domain protein reflects the overall distribution of domains, which may change due to domain rearrangements.

For multi-domain proteins, you should calculate RMSD and RMSF for each domain separately. This requires defining the residue ranges for each domain and performing the analysis on each range independently. You should also calculate the inter-domain distance or angle to assess domain movements.

For large complexes, such as protein-protein or protein-nucleic acid complexes, you should analyze each component separately and together. The overall complex RMSD may be high due to relative movements of the components, even if each component is internally stable. Calculating component-specific RMSD values helps distinguish internal stability from inter-component movements.

## Observations and Measurements in Trajectory Analysis

### Interpreting RMSD Plots for Convergence Assessment

The RMSD plot is the first diagnostic you should examine after a simulation. A well-behaved simulation shows an initial rise in RMSD as the protein relaxes from the starting structure, followed by a plateau where RMSD fluctuates around a constant value. The plateau indicates that the simulation has reached equilibrium and that the protein is sampling conformations around a stable state.

The time required to reach the plateau depends on the system and the starting structure. A homology model may require more time to relax than a crystal structure, because the model may contain strained regions that need to adjust. A protein-ligand complex may require additional time for the ligand to settle into its binding pose.

If the RMSD plot shows a continuous upward trend without a plateau, the simulation may not have equilibrated, or the protein may be undergoing a conformational transition. In this case, you should extend the simulation or examine the trajectory visually to determine what is happening. A sudden jump in RMSD may indicate a conformational transition, such as a loop opening or a domain movement.

### Identifying Flexible Regions from RMSF Profiles

The RMSF profile provides a residue-by-residue view of flexibility. When interpreting RMSF values, you should compare them to the protein's secondary structure and functional annotations. High RMSF values in loop regions are expected and do not necessarily indicate problems. High RMSF values in secondary structure elements may indicate that the structure is unstable or that the simulation conditions are not appropriate.

Terminal regions almost always show high RMSF values because they are not constrained by tertiary interactions. You should not interpret high terminal RMSF values as evidence of instability unless the terminal region is known to be structured. Similarly, surface loops may show high RMSF values because they are exposed to solvent and have fewer stabilizing interactions.

For protein-ligand complexes, comparing RMSF values between the apo and holo simulations can reveal the effect of ligand binding on protein flexibility. A decrease in RMSF at the binding site upon ligand binding indicates that the ligand stabilizes the binding site. An increase in RMSF elsewhere may indicate allosteric effects.

### Using Rg to Detect Unfolding or Compaction Events

The Rg plot shows how the protein's compactness changes over time. A stable protein shows Rg values that fluctuate around a constant value. A gradual increase in Rg indicates that the protein is expanding, which may be a sign of unfolding. A sudden decrease in Rg may indicate a collapse or compaction event.

For intrinsically disordered proteins or proteins that undergo large conformational changes, Rg may show significant variation over time. In these cases, you should examine the trajectory visually to understand the conformational changes that underlie the Rg variations.

Rg is also useful for comparing the compactness of different protein states. For example, you can compare the Rg of a protein in its apo state with its Rg in complex with a ligand to determine whether ligand binding affects the protein's overall compactness.

## Records and Measurements for Reproducible Analysis

### Documenting Analysis Parameters

Reproducible trajectory analysis requires careful documentation of all analysis parameters. For each analysis, you should record the software version, the reference structure used, the atoms included in the calculation, the superposition method, and the frame selection. This information allows other researchers to reproduce your analysis and verify your results.

Many analysis tools produce log files that record the analysis parameters. You should save these log files along with your analysis scripts and output files. If you use command-line tools, you should save the exact commands used for each analysis. If you use Python libraries, you should save the analysis scripts.

The documentation should also include information about the simulation itself, including the simulation software version, the force field, the water model, the temperature and pressure coupling methods, and the simulation length. This information is essential for interpreting the trajectory analysis results and for comparing your results with other studies.

### Storing and Organizing Analysis Output

Trajectory analysis produces multiple output files, including RMSD data files, RMSF data files, Rg data files, and plot images. You should organize these files in a consistent directory structure that makes it easy to find and compare results across different simulations.

A suggested directory structure includes separate folders for each simulation, with subfolders for raw trajectory files, analysis scripts, analysis output, and plots. You should also maintain a master spreadsheet or notebook that records the key results from each analysis, including the average RMSD, the maximum RMSF, and the average Rg.

Version control is important for analysis scripts and documentation. Using a version control system, such as Git, allows you to track changes to your analysis scripts and revert to previous versions if needed. The Carpentries lessons provide foundational training in version control and reproducible computing practices that are directly applicable to trajectory analysis workflows.

### Quality Checks for Analysis Results

Before interpreting your analysis results, you should perform several quality checks to ensure that the results are valid. First, verify that the RMSD values are in the expected range for your system. Values that are unexpectedly high or low may indicate problems with the analysis, such as incorrect superposition or reference structure selection.

Second, verify that the RMSF values are consistent with the protein's known structural features. For example, residues in secondary structure elements should show lower RMSF values than residues in loops. If this pattern is not observed, the analysis may be incorrect.

Third, verify that the Rg values are consistent with the protein's expected size and shape. A globular protein of a given molecular weight has a predictable Rg range. Values outside this range may indicate unfolding or analysis errors.

## Common Failure Patterns in Trajectory Analysis

### Failure to Remove Equilibration Frames

One of the most common errors in trajectory analysis is including equilibration frames in the analysis. The early frames of a simulation often show large structural changes as the system relaxes from the initial conditions. Including these frames in RMSD or RMSF calculations inflates the values and may lead to incorrect conclusions about stability.

The solution is to identify the equilibration period and exclude it from analysis. You can determine the equilibration period by plotting RMSD against time and identifying the point where the RMSD values stabilize. Frames before this point should be excluded from quantitative analysis.

### Incorrect Superposition Leading to Inflated RMSD

Another common error is calculating RMSD without proper superposition. If the structures are not aligned before RMSD calculation, the values include contributions from overall rotation and translation, which inflate the values and obscure internal structural changes.

The solution is to ensure that your analysis tool performs superposition before RMSD calculation. Most tools do this automatically, but you should verify the settings. For RMSF calculation, superposition onto the average structure is essential for removing overall motion and revealing per-residue fluctuations.

### Using Incompatible Topology and Trajectory Files

Using a topology file that does not match the trajectory file produces meaningless results. This can happen if you accidentally use a topology from a different simulation or if the atom ordering in the topology does not match the trajectory.

The solution is to verify that the topology and trajectory files correspond to the same system before analysis. Most analysis tools will produce an error if the files are incompatible, but some may silently produce incorrect results. You should always check the number of atoms in the topology and trajectory files to ensure they match.

### Misinterpreting RMSD Values Without Context

RMSD values are only meaningful in context. An RMSD value of 3 angstroms may indicate a stable simulation for a large flexible protein but may indicate instability for a small rigid protein. Similarly, a high RMSD value may reflect a genuine conformational change instead of instability.

The solution is to interpret RMSD values in the context of your system and your research question. Compare your RMSD values with those reported in similar studies and consider whether the observed deviations are biologically meaningful. Visual inspection of the trajectory can help you determine whether high RMSD values reflect genuine conformational changes or artifacts.

## Limitations of RMSD, RMSF, and Rg Analysis

### Statistical Nature of the Metrics

RMSD, RMSF, and Rg are statistical measures that summarize complex conformational data into single numbers. This statistical nature means that they can obscure important details of the protein's behavior. For example, two different conformational states may have the same average RMSD, even though they represent very different structures.

The limitations of these metrics have led to the development of complementary analysis methods. Trajectory maps, for example, provide a novel method to analyze and visualize protein simulation courses by plotting the protein's backbone movements during the simulation as a heatmap. This approach provides new tools to directly visualize protein behavior over time, compare multiple simulations, and complement established methods. A user-friendly Python application has been developed for this purpose, alongside detailed documentation for easy usage and implementation.

### Sensitivity to Reference Structure Choice

RMSD values are sensitive to the choice of reference structure. Using the first frame as a reference produces different values than using the average structure. This sensitivity means that RMSD values from different studies may not be directly comparable if the studies used different reference structures.

To address this limitation, you should clearly report the reference structure used in your analysis and use consistent reference choices when comparing multiple simulations. You should also consider reporting both the RMSD relative to the starting structure and the RMSD relative to the average structure, as these provide complementary information.

### Inability to Capture Conformational Transitions

RMSD, RMSF, and Rg are equilibrium properties that describe the average behavior of the system. They are not well suited for capturing rare conformational transitions or for describing the pathways between different conformational states. A protein that undergoes a conformational transition during the simulation may show elevated RMSD values, but the metric does not describe the nature of the transition.

For studying conformational transitions, you may need to use additional analysis methods, such as principal component analysis or clustering. These methods can identify the major conformational states sampled during the simulation and the transitions between them. Essential dynamics analysis, which captures principal component analysis and solvent accessible surface area, can reveal the biological motions that correspond to the simulation frames.

## Safety and Regulatory Context for Computational Studies

### Responsible Use of Computational Predictions

Molecular dynamics simulations and trajectory analyses are computational predictions that require experimental validation. The results of these analyses should not be treated as definitive evidence of biological behavior. Instead, they should be used to generate hypotheses that can be tested experimentally.

In drug discovery contexts, computational predictions of protein-ligand stability should be followed by experimental binding assays and functional studies. The trajectory analysis can identify promising candidates, but the biological activity must be confirmed experimentally. Studies that combine molecular docking and molecular dynamics simulations with experimental validation provide the strongest evidence for therapeutic potential.

### Data Management and Reproducibility Requirements

Many funding agencies and journals require that computational studies be reproducible. This means that the simulation parameters, analysis scripts, and raw data must be made available to other researchers. Public data repositories and workflow platforms support this requirement.

The Galaxy Training Network provides accessible workflow training and analysis tutorials that emphasize reproducibility in bioinformatics analyses. The nf-core documentation describes community pipeline standards for reproducible workflow configuration and usage. Following these standards ensures that your trajectory analysis can be reproduced by other researchers.

### Professional Escalation Criteria

If your trajectory analysis produces unexpected results, you should consider whether the issue lies in the simulation or the analysis. The following situations warrant professional escalation:

If your RMSD values continue to rise without plateauing after an extended simulation, the simulation may not be stable. You should consult with experienced molecular dynamics practitioners to diagnose the issue. Possible causes include incorrect simulation parameters, poor starting structure quality, or force field issues.

If your RMSF values show extreme fluctuations that do not match the protein's known structural features, the analysis may be incorrect. You should verify the analysis parameters and consider whether the trajectory file is corrupted or the topology is incorrect.

If your Rg values indicate unfolding but the trajectory visualization does not show unfolding, the analysis may be incorrect. You should verify the Rg calculation and consider whether the trajectory file contains artifacts.

## Practical Implementation Steps for Your First Trajectory Analysis

### Step 1: Verify Your Simulation Output

Before beginning analysis, verify that your simulation completed successfully and that the trajectory file is complete. Check the simulation log file for errors and verify that the trajectory file contains the expected number of frames. If the simulation crashed or the trajectory is incomplete, you may need to rerun the simulation.

### Step 2: Load Your Trajectory and Topology

Load your trajectory and topology files into your chosen analysis tool. Verify that the number of atoms in the topology matches the number of atoms in the trajectory. If you are using GROMACS, you will need the structure file and the trajectory file. If you are using MDAnalysis, you will need the topology file and the trajectory file.

### Step 3: Perform Superposition and Select the Reference Structure

Select the reference structure for your analysis. For most analyses, the first frame or the energy-minimized structure is appropriate. Perform superposition on the protein backbone atoms to remove overall rotation and translation. Verify that the superposition is correct by visualizing a few frames before and after superposition.

### Step 4: Compute and Plot RMSD

Compute RMSD for the protein backbone and plot the values against time. Examine the plot for the equilibration period and the plateau region. If the RMSD values do not plateau, consider whether the simulation needs to be extended or whether the system is undergoing a conformational transition.

### Step 5: Compute and Plot RMSF

Compute RMSF for each residue and plot the values against residue number. Compare the RMSF profile with the protein's secondary structure and functional annotations. Identify the most flexible and most rigid regions and consider whether these patterns are consistent with the protein's known behavior.

### Step 6: Compute and Plot Radius of Gyration

Compute Rg for each frame and plot the values against time. Examine the plot for stability and identify any significant changes in compactness. Compare the average Rg with the expected value for a protein of your system's size and shape.

### Step 7: Document Your Analysis

Record all analysis parameters, including the software version, reference structure, superposition method, and frame selection. Save your analysis scripts and output files. Create a summary of your results, including the average RMSD, the residues with the highest RMSF, and the average Rg.

## Frequently Asked Questions

### What is the difference between RMSD and RMSF?

RMSD measures the deviation of atomic positions from a reference structure over time, providing a global view of structural stability. RMSF measures the fluctuation of each residue around its average position, providing a local view of residue-level flexibility. RMSD is plotted against time, while RMSF is plotted against residue number. RMSD tells you whether the protein remained close to its starting structure, while RMSF tells you which regions moved the most.

### How do I choose the reference structure for RMSD calculation?

The choice of reference structure depends on your research question. Using the first frame as a reference shows how far the protein has moved from its initial conformation and is useful for assessing equilibration. Using the average structure as a reference shows fluctuations around the central structure and is more appropriate for assessing equilibrium dynamics. For comparing multiple simulations, use a common reference, such as the crystal structure or the energy-minimized structure.

### Why is superposition necessary before RMSD calculation?

Superposition removes overall rotation and translation between frames, so that RMSD reflects only internal structural changes. Without superposition, RMSD values include contributions from the protein's overall motion, which inflates the values and obscures internal changes. Most analysis tools perform superposition automatically, but you should verify that the settings match your analysis needs.

### How long should the equilibration period be before analysis?

The equilibration period depends on the system size, the starting structure quality, and the simulation parameters. For a well-prepared system, equilibration may take only a few nanoseconds, while for a poorly prepared system or a large complex, it may take tens of nanoseconds. You can assess equilibration by plotting RMSD against time and identifying the point where the values stabilize.

### What RMSD value indicates a stable simulation?

There is no universal RMSD threshold for stability. The expected RMSD depends on the protein size, flexibility, and the reference structure used. For a globular protein, backbone RMSD values of 1 to 3 angstroms are typical for stable simulations. Values above 5 angstroms may indicate significant conformational changes or instability. You should compare your RMSD values with those reported in similar studies.

### How do I interpret high RMSF values in my protein?

High RMSF values indicate flexible regions. Terminal regions and surface loops typically show high RMSF values because they are less constrained by tertiary interactions. High RMSF values in secondary structure elements may indicate instability. Compare your RMSF profile with the protein's secondary structure and functional annotations to interpret the results in context.

### What does the radius of gyration tell me about my protein?

The radius of gyration measures the distribution of atoms around the center of mass, indicating the protein's compactness. A stable globular protein shows Rg values that fluctuate around a constant value. A gradual increase in Rg may indicate unfolding, while a decrease may indicate compaction. Rg is useful for assessing whether the protein maintains its folded state during simulation.

### Can I compare RMSD values between different simulations?

You can compare RMSD values between simulations if you use the same reference structure and analysis parameters for all simulations. Using a common reference, such as the crystal structure, allows direct comparison. Differences in RMSD between simulations can reveal differences in stability, such as the effect of ligand binding or mutation on protein stability.

## Related Bioinformatics Guides

- [GROMACS Molecular Dynamics: Setting Up, Simulating, and Analyzing Protein-Water Systems](/knowledge/bioinformatics/gromacs-molecular-dynamics-simulation-protocols)
- [Spike Protein Glycan Shield Evolution: Molecular Dynamics Simulation of Immune Evasion in Emerging Coronaviruses](/knowledge/bioinformatics/spike-protein-glycan-shield-evolution-molecular-dynamics-coronaviruses)
- [Volcano Plot Proteomics: How to Create and Interpret Them Effectively](/knowledge/bioinformatics/volcano-plot-proteomics-how-to-create-and-interpret-them-effectively)
- [Protein Language Models in Bioinformatics: A Practical Guide to Selection and Application](/knowledge/bioinformatics/protein-language-models-in-bioinformatics-a-practical-guide-to-selection-and-application)
- [Single-Cell RNA Sequencing Quality Control: A Practical Guide to Filtering and Metrics](/knowledge/bioinformatics/single-cell-rna-sequencing-quality-control-a-practical-guide-to-filtering-and-metrics)

## References and Further Reading

- [NCBI Data Resources](https://www.ncbi.nlm.nih.gov/). National Center for Biotechnology Information.
- [EMBL-EBI Training](https://www.ebi.ac.uk/training). European Bioinformatics Institute.
- [Bioconductor](https://bioconductor.org/). Bioconductor Project.
- [Galaxy Training Network](https://training.galaxyproject.org/). Galaxy Project.
- [nf-core Documentation](https://nf-co.re/docs). nf-core.
- [The Carpentries Lessons](https://carpentries.org/lessons). The Carpentries.
- [Trajectory maps: molecular dynamics visualization and analysis.](https://pubmed.ncbi.nlm.nih.gov/38226394). NAR genomics and bioinformatics, 2024.
- [Molecular modeling, molecular dynamics simulation, and essential dynamics analysis of grancalcin: An upregulated biomarker in experimental autoimmune encephalomyelitis mice.](https://pubmed.ncbi.nlm.nih.gov/36340004). Heliyon, 2022.
- [Exploring allosteric hits of the NS2B-NS3 protease of DENV2 by structure-guided screening.](https://pubmed.ncbi.nlm.nih.gov/37141792). Computational biology and chemistry, 2023.
- [Targeting cadherins with colicins: Molecular docking and dynamics reveal disruptive potential in cancer metastasis.](https://pubmed.ncbi.nlm.nih.gov/40578095). Computational biology and chemistry, 2025.
- [Pharmacokinetic, docking, and DFT analyses reveal Moringa oleifera phytochemicals as inhibitors of HIF-1α/VEGF/GLUT1 signaling pathway in breast cancer.](https://pubmed.ncbi.nlm.nih.gov/41125614). Scientific reports, 2025.

> This article is educational and does not replace validated analysis plans, institutional policy, clinical interpretation, or specialist review.