Geometric Analysis (bioSkills)

A Claude Code skill for taking careful measurements on a static structure: how far apart two atoms are, what the backbone dihedrals do, how much surface is buried, and what an RMSD number actually means.


Type	Claude Skill
Supplier	GPTomics bioSkills (community OSS, MIT)
Availability	GA — part of the bioSkills collection
Pricing	Free / OSS (MIT) — Biopython (BSD-style) and the DSSP binary are installed separately
Capabilities	Read/Write — Claude runs the skill’s workflow locally (Bash/Python), not as an MCP tool
Verified	works · 2026-08-10
Security	cleared · 2026-08-10 — MIT, provenance matches, Biopython BSD, DSSP now Boost/BSD-2

How to install

bioSkills is not an npm package — skills are plain markdown/code read directly by the agent. Clone the repo, then either run the installer for the whole category or copy the single skill directory.

Claude Code — clone and install via the bundled script:
```
git clone https://github.com/GPTomics/bioSkills
cd bioSkills
./install-claude.sh --categories "structural-biology"
```
The installer copies matching skills into ~/.claude/skills/ (default target). Use ./install-claude.sh --list to preview the skills first.
Claude Code / other agents — copy just this one skill:
```
cp -r bioSkills/structural-biology/geometric-analysis ~/.claude/skills/
```
(run from inside your clone — the previous step left you in bioSkills/; otherwise replace bioSkills/ with the absolute path of your clone, e.g. /Users/you/repos/bioSkills).
Prerequisites:
```
pip install "biopython>=1.83" "numpy>=1.26"
conda install -c conda-forge -c bioconda dssp
```
The DSSP v4 binary (mkdssp) is only needed for secondary-structure and some SASA workflows; the distance, dihedral and superposition paths are pure Biopython.

What it does

Parsing and selection — PDBParser into a structure, then atom/residue filtering; the hetflag test (residue.id[0] == ' ') is used to drop waters and heteroatoms before measuring, which is where most silently-wrong numbers come from.
Distances — interatomic distances and full distance matrices.
Angles — bond angles and dihedrals, including backbone phi/psi (Ramachandran) and side-chain chi.
Superposition and RMSD — Kabsch/SVD alignment via Superimposer, or QCPSuperimposer when speed matters.
Bulk descriptors — center of mass, radius of gyration.
Surface — solvent-accessible surface area, reported alongside the probe radius used.
Reporting discipline — every metric is reported with its context: which atoms were selected, which probe radius, which algorithm.

Reference values the skill treats as rules of thumb rather than laws: relative SASA < 0.20 as a common buried-residue heuristic, a rigid-core RMSD cutoff around 2.0 Å, TM-score > 0.5 for “same fold” (length-normalized and asymmetric), and Ramachandran outliers in disallowed regions read as refinement errors rather than biology.

Primary use cases: measuring a specific contact or dihedral to support a mechanistic claim, comparing two conformations of the same protein, quantifying burial of a residue of interest.

Notes

The headline warning is worth internalizing: RMSD is not a property of a structure. It depends jointly on the superposition and the atom selection, and a global all-atom RMSD is dominated by the worst-fitting atoms — it will hide a near-perfect core behind a couple of flexible loops. Related trap: Superimposer requires equal-length, ordered atom lists and does not solve the correspondence problem; feeding it mismatched lists produces a number rather than an error.

This skill measures static structures. For the same quantities computed over a molecular-dynamics trajectory, use MDTraj or MDAnalysis instead; for fold-level comparison across a database rather than a pairwise measurement, see Foldseek. Within the bioSkills structural set it sits downstream of Structure Preparation and alongside Structure Validation (which judges reliability) and Interface Analysis (which applies the same SASA machinery to buried surface between chains); Biopython is the underlying library as its own catalog entry.

Upstream skill front-matter name is bio-structural-biology-geometric-analysis; upstream directory structural-biology/geometric-analysis.

Sources

Installed this tool?

Share feedback — install path, OS, errors, workarounds. The form opens with this tool pre-selected and a link back to this page.