PandaDock-GNN Overview
======================

PandaDock-GNN is an SE(3)-equivariant Graph Neural Network scoring function
for protein-ligand binding affinity prediction.

Key Features
------------

**SE(3)-Equivariance**
   The model produces identical predictions regardless of the rotation or
   translation of the input complex. This is achieved through E(n)-equivariant
   graph neural network (EGNN) layers.

**Heterogeneous Graph Representation**
   Protein and ligand atoms are represented as separate node types in a
   heterogeneous graph, allowing the model to learn distinct representations
   for each.

**Multi-Task Learning**
   The model jointly predicts:

   - **pEC50**: Binding affinity (regression)
   - **Activity**: Binary classification (active/inactive)

Architecture
------------

.. code-block:: text

   Input: Protein-Ligand Complex
          │
          ├─ Protein Atoms → Node Features (56 dims)
          ├─ Ligand Atoms  → Node Features (56 dims)
          └─ Interactions  → Edge Features (23 dims)
          │
          ├─ Node Encoders (separate for protein/ligand)
          │
          ├─ EGNN Layers × 6 (SE(3)-equivariant message passing)
          │   - Update node features
          │   - Update coordinates (equivariant)
          │
          ├─ Attention Pooling
          │   - Protein graph → embedding
          │   - Ligand graph  → embedding
          │
          └─ Prediction Heads
              ├─ Affinity → pEC50
              └─ Activity → probability

Node Features (56 dimensions)
-----------------------------

* Element type one-hot (10 dims)
* SYBYL atom type one-hot (16 dims)
* Partial charge (1 dim)
* Hybridization one-hot (4 dims)
* Aromaticity flag (1 dim)
* H-bond donor/acceptor (2 dims)
* Ring membership (1 dim)
* Residue type one-hot (20 dims, protein only)
* Backbone flag (1 dim, protein only)

Edge Features (23 dimensions)
-----------------------------

* Distance (1 dim)
* Gaussian RBF distance encoding (16 dims)
* Bond type one-hot (4 dims)
* Interaction type flags (2 dims)

Benchmark Performance
---------------------

Tested on ULVSH dataset (942 compounds, 10 protein targets):

+------------------+-------------+
| Metric           | Value       |
+==================+=============+
| Pearson R        | 0.67        |
+------------------+-------------+
| Spearman ρ       | 0.42        |
+------------------+-------------+
| RMSE             | 0.32        |
+------------------+-------------+
| MAE              | 0.12        |
+------------------+-------------+

PandaDock-GNN **outperforms all 8 baseline methods** including:
VM2, MMPBSA, MMGBSA, Gnina, Hyde, DeltaVina, GFN-FF, and PM6.

Usage
-----

**Training:**

.. code-block:: bash

   pandadock gnn train -d ULVSH/ -o models/ --epochs 100

**Prediction:**

.. code-block:: bash

   pandadock gnn predict -m model.pt -p protein.mol2 -l ligand.mol2

**Hybrid Docking (Recommended):**

.. code-block:: bash

   pandadock hybrid -r protein.pdb -l ligand.sdf \
                    --center 10 20 30 --box 20 20 20 \
                    -m model.pt

References
----------

The EGNN architecture is based on:

   Satorras, V. G., Hoogeboom, E., & Welling, M. (2021).
   E(n) Equivariant Graph Neural Networks.
   *International Conference on Machine Learning (ICML)*.