SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2302.12177 · ICML · 2024

EquiPocket: an E(3)-Equivariant Geometric Graph Neural Network for Ligand Binding Site Prediction

Ye Yuan, Zhewei Wei, Chongxuan Li, Yang Zhang, Wenbing Huang

arXiv · PDF · Open in the Atlas

Code that ran

8 of 11 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (2 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 5 have no contract verdict; 3 failures of our instrument; 3 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2302.12177.json · badge: /api/ran/2302.12177.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
fengyuewuya/EquiPocket canonical (classifier label)
classifier basis: classifier_2025
2 ran · 0 no recorded run
ml-jku/vnegnn community
Papers with Code archive
3 ran · 2 no recorded run
fengyuewuya/equipocket community
Papers with Code archive
3 ran · 1 no recorded run
FunctionStatusWhere it lives
Baseline_Models Ran
ran on a synthesized input · constructed an object; no result was computed
fengyuewuya/equipocket/models/EquiPocket.py
code served (permissive licence) · get_code("ff0fdc7478be4958")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, Baseline_Models: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
from torch import nn

torch.manual_seed(0)

ARGS = {
    "atom_channels": 16,
    "bond_channels": 16,
    "out_features": 32,
    "gat_depth": 2,
    "gcn_depth": 0,
    "gin_depth": 0,
    "gcn2_depth": 0,
    "schnet_depth": 0,
    "dimenet_depth": 0,
    "egnn_depth": 0,
    "cutoff": 5,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch, torch_geometric — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Baseline_Models — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a Baseline_Models output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
embed_atom_chem Ran
ran on a synthesized input · constructed an object; no result was computed
fengyuewuya/equipocket/models/EquiPocket.py
code served (permissive licence) · get_code("4d6eda2bb3017c97")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned an instance of its own class, embed_atom_chem: the run constructed the object and did not call it, so no result was computed. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
from torch.nn import Embedding, Sequential, Linear

torch.manual_seed(0)

# The function is embed_atom_chem.__init__, which takes:
# - atom_channels (default 16)
# - formal_charge_channels (default 16)
# - chiral_tag_channels (default 16)
# - aromatic_channels (default 16)
# - ring_channels (default 16)

# These are all just integer channel dimensions, so we can use the defaults
# or provide custom values. Let's use some distinct values for testing.

ARGS = {
    "atom_channels": 16,
    "formal_charge_channels": 16,
    "chiral_tag_channels": 16,
    "aromatic_channels": 16,
    "ring_channels": 16,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type embed_atom_chem — the run constructed an instance of its own class and did not call it; no shape is stated — the instrument coerces any output to an array; for a embed_atom_chem output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
embed_atom_chem Ran
ran on a synthesized input
ml-jku/vnegnn/src/models/equipocket/equipocket.py
code served (permissive licence) · get_code("9c50b31d34f4d0e4")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, ndim 2, dtype family float, shape (2, 82) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
from torch.nn import Embedding

INIT_ARGS = {
    'atom_channels': 16,
    'formal_charge_channels': 16,
    'chiral_tag_channels': 16,
    'aromatic_channels': 16,
    'ring_channels': 16,
}

ARGS = {
    'x': torch.tensor([
        [5, 2, 1, 0, 3, 0.1, 0.2],
        [8, 1, 2, 1, 2, 0.3, 0.4],
    ], dtype=torch.float32),
}

if __name__ == '__main__':
    from torch.nn import Embedding
    
    class embed_atom_chem(torch.nn.Module):

        def __init__(self, atom_channels=16, formal_charge_channels=16, chiral_tag_channels=16, aromatic_channels=16, ring_channels=16):
            super(embed_atom_chem, self).__init__()
            self.embed_atom_channels = Embedding(100, atom_channels)
            self.embed_charge_channels = Embedding(10, formal_charge_channels)
            self.embed_chiral_tag = Embedding(10, chiral_tag_channels)
            self.embed_aromatic_channels = Embedding(10, aromatic_channels)
            self.embed_ring_channels = Embedding(10, ring_channels)

        def forward(self, x):
            embed_atom_channels_data = self.embed_atom_channels(x[:, 0].long())
            embed_charge_channels_data = self.embed_charge_channels(x[:, 1].long())
            embed_chiral_tag_data = self.embed_chiral_tag(x[:, 2].long())
            embed_aromatic_channels_data = self.embed_aromatic_channels(x[:, 3].long())
            embed_ring_channels_data = self.embed_ring_channels(x[:, 4].long())
            all_atom_embed = torch.cat([embed_atom_channels_data, embed_charge_channels_data, embed_chiral_tag_data, embed_aromatic_channels_data, embed_ring_channels_data, x[:, 5:]], axis=1)
            return all_atom_embed

    model = embed_atom_chem(**INIT_ARGS)
    output = model(**ARGS)
    print(output.shape)
    print(output)
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (2, 82) — shape (2, 82) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
embed_bond_chem Ran
ran on a synthesized input
fengyuewuya/equipocket/models/EquiPocket.py
code served (permissive licence) · get_code("504f6367c0d0675e")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, ndim 2, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
from torch.nn import Embedding, Sequential, Linear

class embed_bond_chem(torch.nn.Module):

    def __init__(self, bond_type_channels=16, bond_ring_channels=16):
        super(embed_bond_chem, self).__init__()
        self.embed_bond_type_channels = Embedding(10, bond_type_channels)
        self.embed_bond_ring_channels = Embedding(10, bond_ring_channels)

    def forward(self, edge_attr):
        bond_type_channels_data = self.embed_bond_type_channels(edge_attr[:, 0].long())
        bond_ring_channels_data = self.embed_bond_ring_channels(edge_attr[:, 1].long())
        all_bond_embed = torch.cat([bond_type_channels_data, bond_ring_channels_data, edge_attr[:, 2:]], axis=1)
        return all_bond_embed


INIT_ARGS = {
    'bond_type_channels': 8,
    'bond_ring_channels': 12,
}

ARGS = {
    'edge_attr': torch.tensor([
        [2, 3, 0.5, 0.7],
        [1, 4, 0.3, 0.9],
    ], dtype=torch.float32),
}
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float — no output shape is stated: observe_signature.py observed this class with its constructor sizes canonicalised where it could, which the node does not record, so its shape is not stated as this setup block's
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
embed_bond_chem Ran
ran on a synthesized input
ml-jku/vnegnn/src/models/equipocket/equipocket.py
code served (permissive licence) · get_code("c4bdd9d067583462")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, class construct-then-call prompt, model not recorded) and returned Tensor, ndim 2, dtype family float, shape (2, 34) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch
from torch.nn import Embedding

class embed_bond_chem(torch.nn.Module):

    def __init__(self, bond_type_channels=16, bond_ring_channels=16):
        super(embed_bond_chem, self).__init__()
        self.embed_bond_type_channels = Embedding(10, bond_type_channels)
        self.embed_bond_ring_channels = Embedding(10, bond_ring_channels)

    def forward(self, edge_attr):
        bond_type_channels_data = self.embed_bond_type_channels(edge_attr[:, 0].long())
        bond_ring_channels_data = self.embed_bond_ring_channels(edge_attr[:, 1].long())
        all_bond_embed = torch.cat([bond_type_channels_data, bond_ring_channels_data, edge_attr[:, 2:]], axis=1)
        return all_bond_embed


INIT_ARGS = {
    'bond_type_channels': 16,
    'bond_ring_channels': 16,
}

ARGS = {
    'edge_attr': torch.tensor([
        [3, 5, 0.1, 0.2],
        [7, 2, 0.3, 0.4],
    ], dtype=torch.float32),
}

if __name__ == '__main__':
    model = embed_bond_chem(**INIT_ARGS)
    output = model(**ARGS)
    print(output.shape)
    print(output)
Setup block drafted by
a language model, class construct-then-call prompt, model not recorded
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (2, 34) — shape (2, 34) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: torch and numpy seeded to 0 by the harness immediately before the class was constructed from INIT_ARGS (fixture_check.py)
get_cutoff_ratio Ran
ran; our contract draft was wrong, not the code
fengyuewuya/EquiPocket/models/EquiPocket.py
code served (permissive licence) · get_code("2cbaa0f8a2777675")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned Tensor, ndim 2, dtype family float, shape (5, 5) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create a small position tensor (num_atoms x coordinates)
# Using shape (5, 3) - 5 atoms in 3D space
pos = torch.randn(5, 3)

# Cutoff distance parameter
cutoff = 2.0

# Surface EGNN depth parameter
surface_egnn_depth = 3

ARGS = {
    "pos": pos,
    "cutoff": cutoff,
    "surface_egnn_depth": surface_egnn_depth,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (5, 5) — shape (5, 5) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_cutoff_ratio Ran
ran; our contract draft was wrong, not the code
ml-jku/vnegnn/src/models/equipocket/equipocket.py
code served (permissive licence) · get_code("a8f5c8b4a0c25fb6")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned Tensor, ndim 2, dtype family float, shape (5, 5) as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create a small position tensor (num_atoms=5, coordinates=3)
pos = torch.randn(5, 3, dtype=torch.float32)

# Define cutoff distance
cutoff = 2.0

# Define surface_egnn_depth
surface_egnn_depth = 3

ARGS = {
    "pos": pos,
    "cutoff": cutoff,
    "surface_egnn_depth": surface_egnn_depth,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 2, dtype family float, shape (5, 5) — shape (5, 5) as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_surface_feature Ran
ran; our contract draft was wrong, not the code
fengyuewuya/EquiPocket/protein_feature.py
code served (permissive licence) · get_code("054c05890392b77d")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned tuple. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# vert_surface: tensor with shape (N, M) where first 3 columns are positions
# The function extracts columns [0, 1, 2] as vert_pos
vert_surface = torch.randn(5, 7)  # 5 vertices, 7 features (first 3 are positions)

# protein_pos: tensor of shape (num_atoms, 3) - atomic positions
protein_pos = torch.randn(4, 3)  # 4 atoms, 3D positions

# mean_protein_pos: tensor of shape (3,) - mean position for centering
mean_protein_pos = torch.randn(3)

ARGS = {
    "vert_surface": vert_surface,
    "protein_pos": protein_pos,
    "mean_protein_pos": mean_protein_pos,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
Baseline_Models Not yet run
unverified (harvested, no recorded run)
ml-jku/vnegnn/src/models/equipocket/equipocket.py
code served (permissive licence) · get_code("e5bd90a5490beb9b")
How it ran

no invocation recorded.

EquiPocket Not yet run
unverified (harvested, no recorded run)
fengyuewuya/equipocket/models/EquiPocket.py
code served (permissive licence) · get_code("99ebb0c6e745f5ec")
How it ran

no invocation recorded.

EquiPocket Not yet run
unverified (harvested, no recorded run)
ml-jku/vnegnn/src/models/equipocket/equipocket.py
code served (permissive licence) · get_code("59f89ee880a6f3a9")
How it ran

no invocation recorded.

Also present in other papers' code

Of 11 samples lifted from this paper's linked repositories, 11 carry a function identity and 0 do not. Of the 9 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 9 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Predicting the binding sites of target proteins plays a fundamental role in drug discovery. Most existing deep-learning methods consider a protein as a 3D image by spatially clustering its atoms into voxels and then feed the voxelized protein into a 3D CNN for prediction. However, the CNNbased methods encounter several critical issues: 1) defective in representing irregular protein structures; 2) sensitive to rotations; 3) insufficient to characterize the protein surface; 4) unaware of protein size shift. To address the above issues, this work proposes EquiPocket, an E(3)-equivariant Graph Neural Network (GNN) for binding site prediction, which comprises three modules: the first one to extract local geometric information for each surface atom, the second one to model both the chemical and spatial structure of protein and the last one to capture the geometry of the surface via equivariant message passing over the surface atoms. We further propose a dense attention output layer to alleviate the effect incurred by variable protein size. Extensive experiments on several representative benchmarks demonstrate the superiority of our framework to the state-of-the-art methods. Related codes can be found at the link https://github.com/ fengyuewuya/EquiPocket.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2302.12177")
get_code_for_paper("2302.12177")
have("2302.12177")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2302.12177.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2302.12177.svg)](https://syntology.ai/paper/2302.12177)

Connect an agent — have() is free.