{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Work with protein interfaces\n",
    "\n",
    "Start from a selected interface ID and rebuild its two-chain coordinate view. Use [Query the release tables](2_query_filter_index.ipynb) to select interfaces and {ref}`Protein interfaces <protein-interface-reference>` for the column definitions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import biotite.structure as struc\n",
    "import pandas as pd\n",
    "\n",
    "from plinder.core import PlinderInterface"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Rebuild one interface\n",
    "\n",
    "`PlinderInterface` expands the two selected assembly-chain instances from the source mmCIF revision recorded by the release."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "interface = PlinderInterface(system_id=\"7cm8__1__1.A--2.A\")\n",
    "interface"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.DataFrame(\n",
    "    {\n",
    "        \"chain\": interface.chains,\n",
    "        \"full_sequence_length\": [\n",
    "            len(interface.sequences[chain]) for chain in interface.chains\n",
    "        ],\n",
    "        \"resolved_atom_count\": [\n",
    "            len(interface.chain_structures[chain]) for chain in interface.chains\n",
    "        ],\n",
    "    }\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`sequences` contains the full deposited polymer sequences; coordinate arrays contain only resolved atoms."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Interface residues and coordinate views\n",
    "\n",
    "The residue masks are defined over the complete two-chain `atom_array`. Their keys match the assembly-chain IDs in `interface_annotations`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "summary = []\n",
    "for chain, mask in interface.interface_residue_masks.items():\n",
    "    atoms = interface.atom_array[mask]\n",
    "    summary.append(\n",
    "        {\n",
    "            \"chain\": chain,\n",
    "            \"interface_residues\": struc.get_residue_count(atoms),\n",
    "            \"interface_atoms\": len(atoms),\n",
    "        }\n",
    "    )\n",
    "pd.DataFrame(summary)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "print(f\"Two-chain atoms: {len(interface.atom_array):,}\")\n",
    "print(f\"Interface atoms: {len(interface.interface_structure):,}\")\n",
    "print(f\"Reconstructed mmCIF: {interface.interface_cif}\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`interface_cif` is a `pathlib.Path`; `atom_array`, the values of `chain_structures`, and `interface_structure` are Biotite `AtomArray` objects. The written mmCIF is self-contained and contains only the selected interface chains."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "name": "python",
   "version": "3.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}