README.md

---
pretty_name: Morph processed molecular datasets
language:
  - en
tags:
  - chemistry
  - molecular-generation
  - 3d
  - qm9
  - geom-drugs
  - chembl
  - morph
  - lmdb
  - pytorch-geometric
---

# Morph data

Ready-to-use processed QM9, GEOM-Drugs, and ChEMBL3D data for
[Morph](https://huggingface.co/maltefranke/morph). Download the original processed
files directly, including train/validation/test splits, ordered vocabularies,
training statistics, and SMILES sidecars. No source-data download or preprocessing
is needed to use these caches with Morph.

This dataset repository is public and supports unauthenticated downloads.
It belongs to the
[Morph collection](https://huggingface.co/collections/maltefranke/morph-6aca03baa0d63e6bb9baea6f).

## Available datasets

Sizes are decimal GB. GEOM counts are conformer records; QM9 and ChEMBL3D counts
are molecule records.

| Directory | Representation / selection | Train | Validation | Test | Size |
| --- | --- | ---: | ---: | ---: | ---: |
| `qm9` | Explicit H; original Morph QM9 splits | 100,000 | 20,000 | 12,947 | 0.326 GB |
| `geom` | GEOM-Drugs, explicit H; at most 72 total atoms | 5,645,654 | 703,830 | 706,568 | 56.170 GB |
| `chembl3d` | Explicit H; processed cache without the 100-atom cap | 1,479,308 | 184,947 | 184,960 | 14.795 GB |
| `chembl3d_max100` | Explicit H; at most 100 total atoms | 1,474,008 | 184,295 | 184,242 | 14.726 GB |
| `chembl3d_max100_implicit_h` | Implicit H; same capped source molecules | 1,474,008 | 184,295 | 184,242 | 9.925 GB |

The ChEMBL3D 100-atom cap counts hydrogens **before removal**. The implicit-H
cache has its own vocabularies/statistics and includes the charge/bond
normalization used for the released implicit-H models.

Total processed data: **95,942,750,644 bytes (95.94 GB)** in 61 files.
Scaffold-decoration and protein-pocket datasets are not included in this release.

## Download

From a Morph checkout, download a selected directory into `data/`:

```bash
# Choose one or more datasets:
hf download maltefranke/morph-data --type dataset --include 'qm9/**' --local-dir data
hf download maltefranke/morph-data --type dataset --include 'geom/**' --local-dir data
hf download maltefranke/morph-data --type dataset --include 'chembl3d_max100/**' --local-dir data
hf download maltefranke/morph-data --type dataset --include 'chembl3d_max100_implicit_h/**' --local-dir data

# Uncapped ChEMBL3D cache:
hf download maltefranke/morph-data --type dataset --include 'chembl3d/**' --local-dir data
```

For example, QM9 will be available as `data/qm9/processed/train.pt`, and GEOM as
`data/geom/processed/train_data.lmdb/data.mdb`.

Download everything, including the manifest and checksums:

```bash
hf download maltefranke/morph-data --type dataset --local-dir data
(cd data && sha256sum -c SHA256SUMS)
```

Or use Python:

```python
from huggingface_hub import snapshot_download

snapshot_download(
    repo_id="maltefranke/morph-data",
    repo_type="dataset",
    allow_patterns=["qm9/**"],
    local_dir="data",
)
```

For a reproducible download, pass a dataset commit hash with `--revision` in the
CLI or `revision=` in Python. These are native Morph caches; load them with Morph's
dataset classes rather than `datasets.load_dataset`.

Use a Morph checkout with support for loading processed QM9 caches without raw
inputs. Older QM9 loaders, including those preserved in historical training
bundles, may still fetch raw inputs during dataset initialization.

## Matching the released checkpoints

| Model subfolder in `maltefranke/morph` | Dataset directory here |
| --- | --- |
| `qm9/explicit_h` | `qm9` |
| `geom/explicit_h` | `geom` |
| `chembl3d/explicit_h` | `chembl3d_max100` |
| `chembl3d/implicit_h` | `chembl3d_max100_implicit_h` |

When using a model's training bundle, point `MORPH_DATASET_ROOT` to the downloaded
dataset directory that contains `processed/`. For example:

```bash
export MORPH_DATASET_ROOT="$PWD/data/chembl3d_max100_implicit_h"
```

Vocabulary/statistics SHA-256 hashes and split-file sizes were checked against
the dataset references in the published training bundles. Those historical
references contain split sizes, not split content hashes. This dataset snapshot
adds full SHA-256 checksums for every distributed data file in `SHA256SUMS` and
`manifest.json`.

## File layout

Each dataset directory contains a `processed/` subdirectory with:

- **QM9:** `train.pt`, `val.pt`, `test.pt`, plus PyG preprocessing metadata.
- **GEOM and ChEMBL3D:** `train_data.lmdb/data.mdb`, `val_data.lmdb/data.mdb`,
  `test_data.lmdb/data.mdb`.
- **All datasets:** `atom_tokens.txt`, `edge_tokens.txt`, `charge_tokens.txt`,
  `distributions.pt`, and `train_smiles.txt`, `val_smiles.txt`, `test_smiles.txt`.
- **ChEMBL3D:** original `preprocessing_summary.json`; the capped variants also
  contain `*_source_indices.npy` mappings to their source-cache records.

LMDB records contain atomic numbers (`z`), coordinates (`pos`), charges,
`edge_index`, `edge_attr`, and SMILES. Coordinates are stored in physical Angstroms;
Morph applies the supplied training statistics when constructing model inputs.
SMILES sidecars contain unique split SMILES and are not per-record index maps.
Keep each cache's split files, vocabularies, and statistics together.

Only the LMDB `data.mdb` files are distributed; the Morph reader uses read-only
LMDB access without lock files. Raw inputs, alternative statistics backups,
scaffold caches, and temporary files are omitted.

## Sources and citation

These are processed snapshots of existing molecular datasets. Upstream sources
used by Morph's preprocessing include the
[corrected QM9 structures](https://zenodo.org/records/15700961), the
[GEOM-Drugs Revisited split files](https://bits.csb.pitt.edu/files/geom_raw/), and
LoQI's ChEMBL3D processed release (Figshare file IDs 62280814, 62280805, 62280802).
Consult the upstream releases for their dataset licenses and citation details.

For Morph, please cite
[Generative Molecular Morphing for Flexible-Size Design via Unbalanced Optimal Transport](https://arxiv.org/abs/2606.07239):

```bibtex
@misc{franke2026generativemolecularmorphingflexiblesize,
  title={Generative Molecular Morphing for Flexible-Size Design via Unbalanced Optimal Transport},
  author={Malte Franke and Stefan P. Schmid and Zarko Ivkovic and Kjell Jorner and Andreas Krause},
  year={2026},
  eprint={2606.07239},
  archivePrefix={arXiv},
  primaryClass={cs.LG},
  url={https://arxiv.org/abs/2606.07239},
}
```