Trajectory stability

Trajectory stability

What it does

Takes a Molecular dynamics trajectory and answers the question a raw trajectory file can't answer on its own: did this structure stay stable over the course of the simulation, and if not, which regions moved the most? Use it right after any Molecular dynamics run to check the result before drawing conclusions from it.

How the analysis is done

Three measurements, computed over every frame of the trajectory:

• RMSD over time — how far the structure's backbone has drifted from where it started, at each point in the simulation. A trajectory that plateaus early and stays flat indicates a stable structure; one that keeps climbing has not settled.
• Per-residue flexibility (RMSF) — which specific residues moved the most, on average, once the overall structure's drift through space is accounted for. This is what identifies a flexible loop or terminus versus a rigid core.
• Radius of gyration — a single number per frame tracking how compact the structure stayed overall; a rising radius of gyration means the structure is unfolding or expanding.

Reading the output

The summary reports overall RMSD/RMSF/radius-of-gyration statistics, plus the five most- and five least-flexible residues by name. The table lists per-residue flexibility for the whole structure — read it as "how much did this residue move, on average, once whole-structure drift is subtracted out."

Input

A trajectory and its reference structure — the two files a Molecular dynamics run produces together. They have to come from the same run: the reference is what gives the trajectory's raw coordinates their atom identities.

Important limitations

Backbone and Cα only. This measures the protein's overall shape and per-residue movement, not side-chain conformational changes.

A short simulation gives a real but incomplete picture. The general pattern — which regions are most flexible — tends to show up quickly, but the precise numbers keep refining the longer a simulation runs. Treat a short trajectory's RMSF magnitudes as a reasonable estimate, not a fully converged one; rerun Molecular dynamics with a longer simulation length if you need tighter numbers.