/mdanalysis-trajectory

name: "mdanalysis-trajectory"
description: "Analyze MD trajectories from GROMACS, AMBER, NAMD, CHARMM, LAMMPS. Reads topology/trajectory into Universe objects; supports RMSD, RMSF, radius of gyration, contact maps, H-bonds, PCA, and custom distance/angle calculations. Use for post-simulation structural analysis; use OpenMM/GROMACS for running simulations."
license: "GPL-2.0"

MDAnalysis — Molecular Dynamics Trajectory Analysis

Overview

MDAnalysis provides a uniform Python interface for reading and analyzing molecular dynamics trajectories regardless of MD engine (GROMACS, AMBER, NAMD, CHARMM, LAMMPS, OpenMM). It represents molecular systems as `Universe` objects containing an `AtomGroup` with positions, velocities, forces, and topology data. Trajectories are iterated frame-by-frame or analyzed in bulk using analysis modules for RMSD, RMSF, radius of gyration, hydrogen bonds, solvent-accessible surface area, and PCA. MDAnalysis integrates with NumPy, pandas, and matplotlib, making it the standard tool for post-simulation structural analysis in computational chemistry and drug discovery.

When to Use

• Computing RMSD and RMSF of protein backbone or specific residue groups after MD simulation
• Analyzing ligand binding stability: pocket RMSD, contact persistence, hydrogen bond occupancy
• Performing principal component analysis (PCA) on trajectory conformations
• Computing solvent-accessible surface area (SASA), radius of gyration, and end-to-end distance
• Extracting representative cluster structures from long MD trajectories for visualization
• Use **GROMACS** or **AMBER** analysis tools (`gmx rms`, `cpptraj`) instead for engine-specific analysis within a HPC pipeline
• Use **OpenMM** or **GROMACS** directly for running MD simulations; MDAnalysis is for post-simulation analysis

Prerequisites

• **Python packages**: `MDAnalysis`, `numpy`, `matplotlib`, `pandas`
• **Input**: topology file (.psf, .prmtop, .gro, .pdb) + trajectory file (.dcd, .trr, .xtc, .nc, .dms)

# Install MDAnalysis
pip install MDAnalysis

# Install with all analysis extras
pip install "MDAnalysis[analysis]"

# Verify
python -c "import MDAnalysis as mda; print(mda.__version__)"
# 2.7.0

Quick Start

import MDAnalysis as mda
import numpy as np

# Load a GROMACS topology + trajectory
u = mda.Universe("protein.gro", "trajectory.xtc")

print(f"Atoms: {u.atoms.n_atoms}")
print(f"Residues: {u.residues.n_residues}")
print(f"Frames: {u.trajectory.n_frames}")
print(f"First frame positions (first 3 atoms):\n{u.atoms.positions[:3]}")

Core API

Module 1: Universe and AtomGroup — Loading and Selecting Atoms

Load trajectories and select atom subsets.

import MDAnalysis as mda

# Load topology + trajectory (GROMACS xtc format)
u = mda.Universe("system.gro", "md_production.xtc")

# AtomGroup selections (CHARMM-style selection language)
protein = u.select_atoms("protein")
backbone = u.select_atoms("backbone")
ca_atoms = u.select_atoms("name CA")
ligand = u.select_atoms("resname LIG")
binding_site = u.select_atoms("protein and around 5.0 resname LIG")

print(f"Protein atoms: {protein.n_atoms}")
print(f"CA atoms: {ca_atoms.n_atoms}")
print(f"Ligand atoms: {ligand.n_atoms}")
print(f"Binding site residues: {binding_site.residues.n_residues}")

# Access atom properties at current frame
print(f"CA positions shape: {ca_atoms.positions.shape}")  # (N, 3)
print(f"Protein mass: {protein.total_mass():.1f} Da")

Module 2: Trajectory Iteration — Per-Frame Analysis

Iterate over trajectory frames for time-series analysis.

import MDAnalysis as mda
import numpy as np

u = mda.Universe("protein.gro", "trajectory.xtc")
backbone = u.select_atoms("backbone")

times = []
rg_values = []

for ts in u.trajectory:
    times.append(u.trajectory.time)
    rg_values.append(backbone.radius_of_gyration())

import pandas as pd
df = pd.DataFrame({"time_ps": times, "Rg_A": rg_values})
print(f"Frames analyzed: {len(df)}")
print(f"Mean Rg: {df['Rg_A'].mean():.2f} Å")
print(f"Rg std: {df['Rg_A'].std():.2f} Å")
df.to_csv("radius_of_gyration.csv", index=False)

Module 3: RMSD Analysis — Structural Drift Over Time

Compute backbone RMSD relative to a reference structure.

import MDAnalysis as mda
from MDAnalysis.analysis import rms
import numpy as np
import matplotlib.pyplot as plt

u = mda.Universe("protein.gro", "trajectory.xtc")

# RMSD of Cα atoms relative to first frame
rmsd = rms.RMSD(u, select="name CA")
rmsd.run()

# Results: frame, time (ps), RMSD (Å)
results = rmsd.results.rmsd
print(f"Mean RMSD: {results[:, 2].mean():.2f} Å")
print(f"Max RMSD: {results[:, 2].max():.2f} Å")

# Plot
fig, ax = plt.subplots(figsize=(8, 4))
ax.plot(results[:, 1] / 1000, results[:, 2], color="steelblue", lw=0.8)
ax.set_xlabel("Time (ns)")
ax.set_ylabel("RMSD (Å)")
ax.set_title("Backbone RMSD")
plt.tight_layout()
plt.savefig("rmsd.png", dpi=150)
print("Saved: rmsd.png")

Module 4: RMSF Analysis — Per-Residue Flexibility

Compute root-mean-square fluctuations to identify flexible regions.

import MDAnalysis as mda
from MDAnalysis.analysis import rms
import numpy as np
import matplotlib.pyplot as plt

u = mda.Universe("protein.gro", "trajectory.xtc")

# RMSF per Cα atom (after aligning trajectory)
ca_atoms = u.select_atoms("name CA")
rmsf_analysis = rms.RMSF(ca_atoms)
rmsf_analysis.run()

rmsf_values = rmsf_analysis.results.rmsf
resids = ca_atoms.resids

print(f"Most flexible residue: {resids[np.argmax(rmsf_values)]} ({rmsf_values.max():.2f} Å)")
print(f"Most rigid residue:    {resids[np.argmin(rmsf_values)]} ({rmsf_values.min():.2f} Å)")

# Plot B-factor-like profile
fig, ax = plt.subplots(figsize=(10, 4))
ax.plot(resids, rmsf_values, color="coral", lw=1)
ax.fill_between(resids, 0, rmsf_values, alpha=0.3, color="coral")
ax.set_xlabel("Residue ID")
ax.set_ylabel("RMSF (Å)")
ax.set_title("Per-residue RMSF")
plt.tight_layout()
plt.savefig("rmsf.png", dpi=150)

M