---
title: PDB quickstart
description: Download a PDB structure by ID or search it by name, review it in the 3D viewer, then chain it straight into DiffDock docking or P2Rank pocket prediction.
tags:
  - pdb
  - quickstart
---

You want an experimentally determined structure for a target and then to put it to work — this walks the fastest path: [download it from the PDB](/docs/pdb) by ID (or search by name), review it in the 3D viewer, and chain it straight into docking or binding-site prediction.

## PDB Quickstart

## Step 1: Download a Structure

In chat, request a specific structure:

```
Download 1M17 from PDB
```

Or search by protein name:

```
Search PDB for EGFR kinase structures with resolution below 2.0 Angstroms
```

## Step 2: Review the Structure

The structure loads in an interactive 3D Mol* viewer on the canvas showing:
- Protein chains and residues
- Bound ligands (if present)
- Resolution and experimental method
- Chain identifiers

## Step 3: Chain into Docking

Send the structure to DiffDock:

```
Dock gefitinib into this EGFR structure
```

The platform uses the downloaded PDB structure as the protein input and fetches the ligand SMILES automatically.

## Step 4: Identify Binding Sites

Alternatively, run P2Rank first:

```
Run P2Rank on 1M17 to identify binding pockets
```

This highlights predicted binding sites on the structure, helping you choose where to dock.

## One request, whole pipeline

Describe the whole goal in one message and the orchestrator chains the tools for you:

```
Download 1M17 from PDB, dock gefitinib with DiffDock, and rescore with MM-GBSA
```

See [How the orchestrator picks tools](/docs/how-orchestration-works) for how the chain is built.

## Read next