SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2401.04082 · 2024

Improved motif-scaffolding with SE(3) flow matching

arXiv · PDF · Open in the Atlas

Code that ran

9 of 9 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 0 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 9 failures of our instrument; 0 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/2401.04082.json · badge: /api/ran/2401.04082.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
microsoft/frame-flow canonical (classifier label)
classifier basis: census
8 ran · 0 no recorded run
microsoft/protein-frame-flow canonical (classifier label)
classifier basis: census
1 ran · 0 no recorded run
FunctionStatusWhere it lives
calc_distogram Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
microsoft/frame-flow/models/utils.py
pointer only (licence: MIT); the stored copy get_code serves, from bleach366/p2dflow, is pointer only (licence: NONE) · get_code("04218cc837a6fff9")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from bleach366/p2dflow, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

create_pad_feats Ran
ran; our contract draft was wrong, not the code
microsoft/frame-flow/motif_scaffolding/save_motif_segments.py
code served (permissive licence) · get_code("e78a932b8a492b7b")
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 dict. 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)

# The function create_pad_feats takes a single parameter: pad_amt (an integer)
# It uses pad_amt to create a tensor of shape (pad_amt, 4, 4)
# So pad_amt should be a positive integer

ARGS = {
    'pad_amt': 2
}
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 dict — a dict output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a dict 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
dataset_creation Ran
ran; our contract draft was wrong, not the code
microsoft/frame-flow/experiments/utils.py
code served (permissive licence) · get_code("970c12943df8215f")
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 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
import torch.nn as nn

torch.manual_seed(0)

class StandInDataset:
    def __init__(self, dataset_cfg, task, is_training):
        self.dataset_cfg = dataset_cfg
        self.task = task
        self.is_training = is_training
        self.data = [
            (torch.randn(4), torch.tensor([0])),
            (torch.randn(4), torch.tensor([1]))
        ]
    
    def __len__(self):
        return len(self.data)
    
    def __getitem__(self, idx):
        return self.data[idx]

dataset_class = StandInDataset

cfg = {
    "num_classes": 2,
    "batch_size": 2,
    "learning_rate": 0.001
}

task = "classification"

ARGS = {
    "dataset_class": dataset_class,
    "cfg": cfg,
    "task": task
}
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 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
flatten_final_dims Ran
ran; our contract draft was wrong, not the code
microsoft/frame-flow/models/ipa_pytorch.py
code served (permissive licence) · get_code("706bd906bbbd5b15")
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 Tensor, ndim 3, dtype family float, shape (2, 4, 15) 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 tensor with shape that allows flattening final dimensions
# Using distinct sizes: batch=2, dim1=4, dim2=3, dim3=5
t = torch.randn(2, 4, 3, 5)

# no_dims specifies how many final dimensions to flatten
no_dims = 2

ARGS = {
    "t": t,
    "no_dims": no_dims,
}
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 Tensor, ndim 3, dtype family float, shape (2, 4, 15) — shape (2, 4, 15) 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_index_embedding Ran
ran; our contract draft was wrong, not the code
microsoft/frame-flow/models/utils.py
pointer only (licence: MIT); the stored copy get_code serves, from bleach366/p2dflow, is pointer only (licence: NONE) · get_code("45bba8bf8e695be0")
How it ran

The body of this sample is withheld: this paper's link allows the body inline (MIT), but get_code serves the stored copy, first harvested from bleach366/p2dflow, and GitHub detected no licence in that repository when it was checked. The setup block that called it is withheld with it, because it is written against that body. source_path and origin say where the code lives, to read under the repository's own terms. A run is recorded.

get_pdb_headers Ran
ran; our contract draft was wrong, not the code
microsoft/protein-frame-flow/openfold/np/protein.py
code served (permissive licence) · get_code("576c6e55a16fd58e")
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 list, ndim 1, 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 dataclasses
from typing import Any, Sequence, Mapping, Optional
import numpy as np

@dataclasses.dataclass(frozen=True)
class Protein:
    """Protein structure representation."""
    atom_positions: np.ndarray
    aatype: np.ndarray
    atom_mask: np.ndarray
    residue_index: np.ndarray
    b_factors: np.ndarray
    chain_index: Optional[np.ndarray] = None
    remark: Optional[str] = None
    parents: Optional[Sequence[str]] = None
    parents_chain_index: Optional[Sequence[int]] = None

# Create a minimal Protein instance
prot = Protein(
    atom_positions=np.random.randn(10, 3).astype(np.float32),
    aatype=np.array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9], dtype=np.int32),
    atom_mask=np.ones(10, dtype=np.float32),
    residue_index=np.arange(10, dtype=np.int32),
    b_factors=np.ones(10, dtype=np.float32),
    chain_index=np.array([0, 0, 0, 1, 1, 1, 2, 2, 2, 2], dtype=np.int32),
    remark="Test protein structure",
    parents=["parent_A", "parent_B", "parent_C"],
    parents_chain_index=[0, 0, 1]
)

ARGS = {
    "prot": prot,
    "chain_id": 0
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type list, ndim 1, dtype family float — a list output has no array shape of its own; none is stated
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_sampled_mask Ran
ran; our contract draft was wrong, not the code
microsoft/frame-flow/experiments/utils.py
code served (permissive licence) · get_code("dd8d6f9f0c7a99d8")
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 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 random
import numpy as np

# Construct arguments for get_sampled_mask function
ARGS = {
    "contigs": "A0-10 B5-15 C,10-20,0",
    "length": [20, 50],
    "rng": np.random.default_rng(seed=42),
    "num_tries": 1000000
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
numpy — 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
get_time_embedding Ran
ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)
microsoft/frame-flow/models/utils.py
code served (permissive licence) · get_code("d64bededcdfc5d77")
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 Tensor, ndim 2, dtype family float, shape (2, 4) 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
import math

torch.manual_seed(0)

# Create valid arguments for get_time_embedding
timesteps = torch.tensor([0, 1], dtype=torch.int64)  # Shape: (2,)
embedding_dim = 4
max_positions = 2000

ARGS = {
    "timesteps": timesteps,
    "embedding_dim": embedding_dim,
    "max_positions": max_positions,
}
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 Tensor, ndim 2, dtype family float, shape (2, 4) — shape (2, 4) 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
permute_final_dims Ran
ran; our contract draft was wrong, not the code
microsoft/frame-flow/models/ipa_pytorch.py
code served (permissive licence) · get_code("c653d7a1bb820cd6")
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 Tensor, ndim 4, dtype family float, shape (2, 3, 5, 4) 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 typing import List

torch.manual_seed(0)

# Create a small tensor with distinct dimensions for probing
# Shape: (2, 3, 4, 5) - batch=2, some intermediate dims, final dims that will be permuted
tensor = torch.randn(2, 3, 4, 5)

# Create a permutation of the final dimensions
# If we want to permute the last 2 dimensions [4, 5], we can reverse them
inds = [1, 0]

ARGS = {
    "tensor": tensor,
    "inds": inds,
}
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 Tensor, ndim 4, dtype family float, shape (2, 3, 5, 4) — shape (2, 3, 5, 4) 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

Also present in other papers' code

Of 9 samples lifted from this paper's linked repositories, 9 carry a function identity and 0 do not. Of the 9 distinct functions with an identity, 6 are also present in other papers whose code we hold (2 with at least one paper dated earlier, 0 the same month, 6 later) and 3 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.

FunctionAlso present inEarliest-datedOthers named
flatten_final_dims 10 other papers
1 dated earlier, 9 later · 1 only as an assembly brick
2305.14749 (2023-05)
gRNAde: Geometric Deep Learning for 3D RNA inverse design
2405.20313 (2024-05)
Sequence-Augmented SE(3)-Flow Matching For Conditional Protein Backbone Generation
2406.00735 (2024-06)
Full-Atom Peptide Design based on Multi-modal Flow Matching
and 7 more not named here
permute_final_dims 10 other papers
1 dated earlier, 9 later · 1 only as an assembly brick
2305.14749 (2023-05)
gRNAde: Geometric Deep Learning for 3D RNA inverse design
2405.20313 (2024-05)
Sequence-Augmented SE(3)-Flow Matching For Conditional Protein Backbone Generation
2406.00735 (2024-06)
Full-Atom Peptide Design based on Multi-modal Flow Matching
and 7 more not named here
get_index_embedding 6 other papers
6 later
2405.20313 (2024-05)
Sequence-Augmented SE(3)-Flow Matching For Conditional Protein Backbone Generation
2406.00735 (2024-06)
Full-Atom Peptide Design based on Multi-modal Flow Matching
2406.03141 (2024-06)
Floating Anchor Diffusion Model for Multi-motif Scaffolding
and 3 more not named here
calc_distogram 4 other papers
4 later
2406.00735 (2024-06)
Full-Atom Peptide Design based on Multi-modal Flow Matching
2411.05238 (2024-11)
Generating Highly Designable Proteins with Geometric Algebra Flow Matching
2411.17196 (2024-11)
P2DFlow: A Protein Ensemble Generative Model with SE(3) Flow Matching
and 1 more not named here
dataset_creation 1 other paper
1 later
2503.05025 (2025-03)
ProtComposer: Compositional Protein Structure Generation with 3D Ellipsoids
get_time_embedding 1 other paper
1 later
2503.05025 (2025-03)
ProtComposer: Compositional Protein Structure Generation with 3D Ellipsoids

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

Protein design often begins with the knowledge of a desired function from a motif which motif-scaffolding aims to construct a functional protein around. Recently, generative models have achieved breakthrough success in designing scaffolds for a range of motifs. However, generated scaffolds tend to lack structural diversity, which can hinder success in wet-lab validation. In this work, we extend FrameFlow, an SE(3) flow matching model for protein backbone generation, to perform motif-scaffolding with two complementary approaches. The first is motif amortization, in which FrameFlow is trained with the motif as input using a data augmentation strategy. The second is motif guidance, which performs scaffolding using an estimate of the conditional score from FrameFlow without additional training. On a benchmark of 24 biologically meaningful motifs, we show our method achieves 2.5 times more designable and unique motif-scaffolds compared to state-of-the-art. Code: https://github.com/microsoft/protein-frame-flow

For agents

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

get_harvested_code_for_paper("2401.04082")
get_code_for_paper("2401.04082")
have("2401.04082")

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

curl https://syntology.ai/api/ran/2401.04082.json

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

[![Syntology run record](https://syntology.ai/api/ran/2401.04082.svg)](https://syntology.ai/paper/2401.04082)

Connect an agent — have() is free.