How does this skill improve upon manual PDB preparation or basic LLM prompting?

The skill automates the removal of crystallization waters, additives, and PEG while extracting specific ligands and calculating precise XYZ docking coordinates, eliminating the manual errors common in text-based PDB preparation.

Can I use my own local PDB files, or does it only work with RCSB database IDs?

Yes, the skill can process both 4-character RCSB PDB IDs for automated downloading and local .pdb files stored on your system.

Does 'pdb prep' handle hydrogen addition and protonation state assignments?

While the skill cleans and isolates structures, it does not currently add hydrogens or assign specific protonation states; we recommend using a tool like Open Babel or MGLTools for these final steps.

What specific outputs are provided for docking simulations?

The skill provides ligand centroids and search box dimensions (size_x, y, z) specifically formatted for seamless integration into AutoDock Vina and Gnina configuration files.

What exactly is included in the purchase of this skill?

A SKILL.md that tells your agent when and how to use it, plus the self-tested pdbprep.py script that does the fetching, cleaning, binding-site detection and box calculation. One file and one script; no account or subscription.

What are the technical requirements to run this skill on my AI agent?

Only the uv package manager. The script uses the Python standard library, so there is no RDKit or other dependency to install. It runs in any agent that can execute shell commands.

PDB prep for docking: receptor, ligand and Vina box

Example session with this skill installed

Prep PDB 1HSG for docking.

Read your context and instructions
Compiled the pdb prep

Summary

Fetched 1HSG (HIV-1 protease with indinavir) from RCSB, cleaned the receptor and extracted the ligand. Script output:

Resolution: 2.0 A
Chains: A (99 residues), B (99 residues)
Ligands (water/additives excluded): MK1 902:B
Missing residues (REMARK 465): 0 | Chain breaks: none
Cleaned receptor: 1HSG_protein.pdb | Ligand: 1HSG_ligand.pdb
Binding site (<=5 A, 33 residues): ARG8:A, LEU23:A, ASP25:A, GLY27:A, ALA28:A,
  ASP29:A, ASP30:A, VAL32:A, ILE47:A, GLY48:A, GLY49:A, ILE50:A, PHE53:A,
  PRO81:A, VAL82:A, ILE84:A, ARG8:B, LEU23:B, ASP25:B, GLY27:B, ALA28:B,
  ASP29:B, ASP30:B, THR31:B, VAL32:B, ILE47:B, GLY48:B, GLY49:B, ILE50:B,
  LEU76:B, PRO81:B, VAL82:B, ILE84:B
Docking box (Vina/Gnina):
  center_x = 13.073   size_x = 30.9
  center_y = 22.467   size_y = 26.2
  center_z = 5.557    size_z = 24.6

What this tells you

The site contains the catalytic

ASP25 on both chains, which is the expected HIV-1 protease active site, so the box is centred correctly. At 2.0 A with no gaps and no missing residues, the structure is clean enough to dock against as-is.

Config, ready to paste

receptor = 1HSG_protein.pdbqt
center_x = 13.073
center_y = 22.467
center_z = 5.557
size_x = 30.9
size_y = 26.2
size_z = 24.6

Next steps

Add hydrogens and set protonation at your target pH (PDB2PQR, Open Babel or Reduce), then convert to PDBQT.
The box spans both chains' sites. For a single site, rerun with --chain A --pad 6 to tighten it.
Use 1HSG_ligand.pdb as the redocking control: an RMSD under 2 A to the crystal pose means the setup is sound.

Connects securely to your tools. The creator never sees your data.

About this skill

Turns a PDB entry into docking-ready inputs in one step.

What it does

Fetch or load: downloads from RCSB by ID, or reads a local .pdb file.
Inspect: reports resolution, chains with residue counts, ligands and HETATM groups (water and ~25 common buffer/ion additives excluded), missing residues from REMARK 465, and chain breaks found from numbering gaps.
Clean: writes a receptor PDB with the first model only, altloc A, and no water, additives or other ligands. --chain and --keep-ligand control what stays.
Split: writes the selected ligand as its own PDB.
Box: lists binding-site residues within 5 A of the ligand and prints a Vina/Gnina box (centre and size, padding adjustable with --pad).
Warn: flags resolution worse than 2.5 A, gaps near the site, box edges over 30 A and multiple ligand candidates.

Why it's trustworthy

All parsing and geometry run in a bundled, self-tested script (scripts/pdbprep.py --selftest); the model never invents coordinates. The script uses only the Python standard library, so there is nothing to install beyond uv.

Requirements

uv, and internet access only when fetching a structure by PDB ID.

How to use

Skills on Agensi run inside your AI agent. Copy the prompt, paste it into your agent, and it connects to Agensi and installs this skill for you. If a payment is needed, your agent shows it to you first.

Reviews

No reviews yet

Be one of the first to try it. Every listed skill passes our trust checks below.

Security scanned

Passed our 8-point scan before listing

Fresh listing

Recently published to Agensi

Money-back guarantee

Doesn't do what its page promises? You get your money back.

PDB prep for docking: receptor, ligand and Vina box