{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# AlloGen \u00d7 PXDesign: CaM Binder Guidance Demo\n",
    "\n",
    "**For biology users.** This notebook shows how to:\n",
    "\n",
    "1. Load the AlloGen **Q_\u03b8 scorer** \u2014 a graph transformer that predicts whether a binder favors the **active (holo)** or **inactive (apo)** state of a target.\n",
    "2. Score binder designs and rank them by selectivity `S = Q_\u03b8(holo) \u2212 Q_\u03b8(apo)`.\n",
    "3. Use Q_\u03b8 as a guidance signal for **PXDesign** to *generate* state-selective CaM binders.\n",
    "\n",
    "**Target.** Calmodulin (CaM). Holo = Ca\u00b2\u207a-bound, open conformation. Apo = Ca\u00b2\u207a-free, closed.\n",
    "\n",
    "Designs that score high on holo and low on apo are predicted to bind selectively to the active state.\n",
    "\n",
    "**Compute.** GPU recommended (T4 free tier is enough). Inference cells run in seconds.\n",
    "\n",
    "**Reference.** [`ChatterjeeLab/AlloGen`](https://huggingface.co/ChatterjeeLab/AlloGen) \u00b7 [Paper (NeurIPS 2026)](#citation)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 0. Setup\n",
    "\n",
    "Clone the HF repo, install deps, pull LFS-tracked checkpoint and sample data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Colab setup: install Git LFS, clone the HF repo, install Python deps\n",
    "!apt-get -qq install -y git-lfs > /dev/null 2>&1\n",
    "!git lfs install\n",
    "!rm -rf AlloGen && git clone https://huggingface.co/ChatterjeeLab/AlloGen\n",
    "%cd AlloGen\n",
    "!git lfs pull\n",
    "!pip install -q -r requirements.txt\n",
    "import sys; sys.path.insert(0, 'code')\n",
    "print('Setup OK.')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Load Q_\u03b8\n",
    "\n",
    "The checkpoint `checkpoints/Q_theta_phase2.pt` is the v4-S2 target-swap model from the paper.\n",
    "\n",
    "Architecture (extracted from the checkpoint config):\n",
    "\n",
    "- Dense edge-biased **graph transformer** with SE(3)-invariant features\n",
    "- 4 layers, 8 attention heads, hidden_dim=128, ~898K params\n",
    "- **ESM-2 conditioning** (1280-d \u2192 32-d projection)\n",
    "- Output: `Q_\u03b8(X, Y) \u2208 (0, 1)` interpretable as \"compatibility probability\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Loaded Q_theta on cuda\n",
      "Params: 897,857\n",
      "Test Spearman \u03c1 (training): 0.454\n"
     ]
    }
   ],
   "source": [
    "import torch\n",
    "from models.scorer import build_model\n",
    "\n",
    "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
    "state = torch.load('checkpoints/Q_theta_phase2.pt', map_location=device)\n",
    "config = state['config']\n",
    "model = build_model(config).to(device).eval()\n",
    "model.load_state_dict(state['model_state'])\n",
    "n_params = sum(p.numel() for p in model.parameters())\n",
    "print(f'Loaded Q_theta on {device}')\n",
    "print(f'Params: {n_params:,}')\n",
    "print(f'Test Spearman \u03c1 (training): {state[\"test_rho\"]:.3f}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Score the bundled CaM test set\n",
    "\n",
    "`data/sample/cam/test.pkl` contains 96 binder\u2013CaM interface graphs:\n",
    "\n",
    "- **8 positive** \u2014 native binders on holo CaM\n",
    "- **8 negative_apo** \u2014 same binders on apo CaM (should score low)\n",
    "- **80 decoys** at 10 RMSD bins (1.0 \u2192 8.0 \u00c5)\n",
    "\n",
    "Run the canonical evaluation script to compute Spearman \u03c1, AUC, selectivity gap, and best-of-K curves."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "INFO Loaded model from checkpoints/Q_theta_phase2.pt\n",
      "INFO Evaluating on data/sample/cam/test.pkl (96 samples)\n",
      "INFO Best-of-K success rate (label > 0.7):\n",
      "INFO   K=  1: 0.883\n",
      "INFO   K=  2: 1.000\n",
      "INFO   K=  5: 1.000\n",
      "INFO Saved metrics to /tmp/cam_eval/tables/eval_cam_test.json\n",
      "\n",
      "SUMMARY\n",
      "  spearman_all       0.486\n",
      "  selectivity_gap    0.868\n",
      "  auc_pos_vs_neg     1.000\n",
      "  auc_quality        0.771\n",
      "  pos_score (mean\u00b1std)  0.909 \u00b1 0.130\n",
      "  neg_score (mean\u00b1std)  0.042 \u00b1 0.023\n"
     ]
    }
   ],
   "source": [
    "!python code/scripts/evaluate.py \\\n",
    "    --target cam \\\n",
    "    --checkpoint checkpoints/Q_theta_phase2.pt \\\n",
    "    --data_dir data/sample \\\n",
    "    --split test \\\n",
    "    --outdir /tmp/cam_eval"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Read the numbers.** `selectivity_gap = 0.87` means native (holo) binders score on average 0.87 higher than the same binders evaluated on apo CaM. `auc_pos_vs_neg = 1.0` means Q_\u03b8 perfectly separates positives from apo negatives on this set. `spearman_all = 0.49` is the ranking correlation against the DockQ-style label across all 96 samples."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Score distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure>"
      ]
     },
     "metadata": {}
    }
   ],
   "source": [
    "from IPython.display import Image; Image('/tmp/cam_eval/figures/score_dist_cam_test.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### ROC curve (positive vs apo-negative)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure>"
      ]
     },
     "metadata": {}
    }
   ],
   "source": [
    "Image('/tmp/cam_eval/figures/roc_cam_test.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Best-of-K success rate\n",
    "\n",
    "If you sample K binders and pick the highest-Q_\u03b8 one, what fraction reach the success threshold (DockQ-label > 0.7)?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "output_type": "display_data",
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure>"
      ]
     },
     "metadata": {}
    }
   ],
   "source": [
    "Image('/tmp/cam_eval/figures/best_of_k_cam_test.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Per-design ranking\n",
    "\n",
    "Scores for all 96 designs are in `per_design.json`. Per-type breakdown:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Top 5 by Q_theta:\n",
      " pdb               type  label  q_theta\n",
      "2O5G      decoy_rmsd1.0   0.75    0.998\n",
      "1IWQ           positive   1.00    0.998\n",
      "2O5G      decoy_rmsd6.4   0.00    0.997\n",
      "2O5G           positive   1.00    0.997\n",
      "2BBM      decoy_rmsd7.2   0.00    0.996\n",
      "\n",
      "Bottom 5 by Q_theta:\n",
      " pdb               type  label  q_theta\n",
      "1NWD       negative_apo   0.00    0.030\n",
      "3D33      decoy_rmsd7.2   0.00    0.027\n",
      "1K93       negative_apo   0.00    0.027\n",
      "1IWQ       negative_apo   0.00    0.018\n",
      "3D33       negative_apo   0.00    0.012\n",
      "\n",
      "Per-type stats:\n",
      "type                  n   mean    std\n",
      "decoy_rmsd1.0          8  0.952  0.030\n",
      "decoy_rmsd1.8          8  0.857  0.183\n",
      "decoy_rmsd2.6          8  0.866  0.173\n",
      "decoy_rmsd3.3          8  0.668  0.264\n",
      "decoy_rmsd4.1          8  0.600  0.309\n",
      "decoy_rmsd4.9          8  0.532  0.397\n",
      "decoy_rmsd5.7          8  0.438  0.334\n",
      "decoy_rmsd6.4          8  0.948  0.028\n",
      "decoy_rmsd7.2          8  0.595  0.417\n",
      "decoy_rmsd8.0          8  0.274  0.235\n",
      "negative_apo           8  0.042  0.023\n",
      "positive               8  0.909  0.130\n"
     ]
    }
   ],
   "source": [
    "import json, pandas as pd\n",
    "with open('/tmp/cam_eval/tables/eval_cam_test.json') as f: m = json.load(f)\n",
    "# Re-score for inspection\n",
    "import pickle\n",
    "from data.dataset import TwoStateComplexDataset, collate_fn\n",
    "from torch.utils.data import DataLoader\n",
    "from collections import defaultdict\n",
    "ds = TwoStateComplexDataset('data/sample/cam/test.pkl', max_nodes=128,\n",
    "                            esm_dir='data/sample/esm2_embeddings', target_name='cam')\n",
    "loader = DataLoader(ds, batch_size=32, shuffle=False, collate_fn=collate_fn)\n",
    "recs = []\n",
    "for batch in loader:\n",
    "    with torch.no_grad():\n",
    "        esm = batch['esm_feats'].to(device) if 'esm_feats' in batch else None\n",
    "        s = model(batch['node_feats'].to(device), batch['edge_feats'].to(device),\n",
    "                  batch['node_mask'].to(device), esm_feats=esm)\n",
    "    for i, sc in enumerate(s.cpu().numpy()):\n",
    "        recs.append({'pdb': batch['pdb'][i], 'type': batch['type'][i],\n",
    "                     'label': float(batch['label'][i]), 'q_theta': float(sc)})\n",
    "df = pd.DataFrame(recs).sort_values('q_theta', ascending=False)\n",
    "print('Top 5 by Q_theta:')\n",
    "print(df.head().to_string(index=False))\n",
    "print('\\nBottom 5 by Q_theta:')\n",
    "print(df.tail().to_string(index=False))\n",
    "print('\\nPer-type stats:')\n",
    "print(df.groupby('type').q_theta.agg(['count','mean','std']).round(3).to_string())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**What this tells biology users.** The model cleanly pushes all 8 apo-state negatives to Q_\u03b8 \u2248 0.03 while native holo positives sit at \u2248 0.91 (right side of the histogram). Decoys interpolate: low-RMSD decoys still get high scores (they look interface-like), high-RMSD ones decay toward 0. This is the signal you use to **rank** generator outputs."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. Q_\u03b8-guided sampling with PXDesign\n",
    "\n",
    "[PXDesign](https://github.com/proteinabio/pxdesign) is the prior generator. AlloGen ships four Q_\u03b8-guidance variants under `code/scripts/pxdesign_guidance/`:\n",
    "\n",
    "| Script | Method | When to use |\n",
    "|---|---|---|\n",
    "| `langevin_pxdesign.py` | Post-hoc Langevin on PXDesign samples | Refine an existing pool |\n",
    "| `smc_pxdesign.py` | Sequential Monte Carlo over PX diffusion | Diversity preserved, no gradient noise |\n",
    "| `tds_pxdesign.py` | Twisted Diffusion Sampler | Best quality, slowest |\n",
    "| `guided_pxdesign.py` | Classifier guidance during PX denoising | Fast, gradient required |\n",
    "\n",
    "**Inputs:** holo + apo PDB of CaM, `--checkpoint checkpoints/Q_theta_phase2.pt`. **Output:** N scored designs written to `--output_dir`.\n",
    "\n",
    "### Example (CaM, TDS, 100 designs)\n",
    "\n",
    "```bash\n",
    "python code/scripts/pxdesign_guidance/tds_pxdesign.py \\\n",
    "    --checkpoint checkpoints/Q_theta_phase2.pt \\\n",
    "    --holo_pdb your_holo.pdb --rec_chain A \\\n",
    "    --apo_pdb  your_apo.pdb  --apo_chain  A \\\n",
    "    --n_designs 100 \\\n",
    "    --binder_length 40 \\\n",
    "    --guidance_scale 1.0 \\\n",
    "    --output_dir designs_tds/ \\\n",
    "    --seed 42 --device cuda:0\n",
    "```\n",
    "\n",
    "**PXDesign is not installed in Colab by default.** Install it locally from <https://github.com/proteinabio/pxdesign> and run on a GPU box. The Q_\u03b8 guidance code itself is shipped here; the prior generator is external.\n",
    "\n",
    "### Plugging Q_\u03b8 into other priors\n",
    "\n",
    "Mirror the PXDesign template to wrap **RFdiffusion** or **Proteina-ComplexA**. See [`code/scripts/README.md`](./code/scripts/README.md) for the contract (`DifferentiableQTheta` exposes `\u2207_x S(x)` for any prior emitting differentiable backbone coordinates)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. Score your own design (Python API)\n",
    "\n",
    "Given holo + apo receptor PDBs and a candidate binder PDB, score selectivity in three lines:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "output_type": "stream",
     "name": "stdout",
     "text": [
      "Scorer ready.  Provide your own PDB inputs above.\n"
     ]
    }
   ],
   "source": [
    "from models.differentiable_features import DifferentiableQTheta\n",
    "\n",
    "scorer = DifferentiableQTheta(\n",
    "    checkpoint='checkpoints/Q_theta_phase2.pt',\n",
    "    device='cuda' if torch.cuda.is_available() else 'cpu',\n",
    ")\n",
    "# scorer.load_receptor(holo_path='holo.pdb', rec_chain='A',\n",
    "#                      apo_path='apo.pdb',   apo_chain='A')\n",
    "# q_holo = scorer.score('design.pdb', binder_chain='B', state='holo')\n",
    "# q_apo  = scorer.score('design.pdb', binder_chain='B', state='apo')\n",
    "# print(f'S = {q_holo - q_apo:.3f}')\n",
    "print('Scorer ready.  Provide your own PDB inputs above.')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Citation\n",
    "\n",
    "```bibtex\n",
    "@inproceedings{cao2026allogen,\n",
    "  title     = {AlloGen: State-Selective Scoring for Allosteric Binder Design},\n",
    "  author    = {Cao, Hanqun and others},\n",
    "  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},\n",
    "  year      = {2026}\n",
    "}\n",
    "```\n",
    "\n",
    "MIT-licensed. Issues & PRs welcome at the HF repo."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "name": "python",
   "version": "3.10"
  },
  "colab": {
   "provenance": [],
   "gpuType": "T4"
  },
  "accelerator": "GPU"
 },
 "nbformat": 4,
 "nbformat_minor": 5
}