---
title: "Molecule benchmarks (QM9, ZINC, MUTAG): tasks and access"
description: "Molecule benchmarks (QM9, ZINC, MUTAG): compare tasks, official access, usage conditions and evaluation. Includes research-paper mention counts and trends."
url: https://scholar4ai.com/datasets/molecule-benchmarks-qm9-zinc-mutag
markdown_url: https://scholar4ai.com/datasets/molecule-benchmarks-qm9-zinc-mutag.md
license: CC BY-NC 4.0 (https://creativecommons.org/licenses/by-nc/4.0/)
share_text: "Molecule benchmarks (QM9, ZINC, MUTAG): compare tasks, official access, usage conditions and evaluation. Includes research-paper mention counts and trends. (as of 2026-11-05)"
data_url: https://scholar4ai.com/data/rankings.json
updated: 2026-11-05T06:02:37+09:00
window: 2026-10-07 to 2026-11-05
source: OpenAlex (CC0)
---

# Molecule benchmarks (QM9, ZINC, MUTAG)

[Browse dataset profiles](https://scholar4ai.com/datasets/directory/page/4.md)

Citation, molecule and other graphs used to compare graph neural networks.

- data category: Graph learning benchmarks
- 0 papers in the last 7 days
- Names searched for: QM9, ZINC250k, MUTAG, MoleculeNet, TUDataset

Most-used data: https://scholar4ai.com/rankings/datasets.md · All Graph learning benchmarks datasets: https://scholar4ai.com/datasets/graph-learning-benchmarks.md

## Tasks, access and usage conditions

Mention counts below measure name matches in tracked papers, not benchmark quality or downloads. A family count can cover multiple datasets; it is not a separate count for each comparison row. Choose a task and version first, preserve evaluation splits, and follow the linked release terms. The site's content license does not replace the datasets' own terms.

*Dataset comparison: tasks, access and usage conditions*

| Dataset / version | Task and evaluation | Access and preparation | Usage conditions |
|---|---|---|---|
| QM9 | Molecular-property regression from computed small-molecule geometries and properties. Evaluation: Report per-target MAE with units and split; quantum-chemistry targets are computed labels, not clinical outcomes. | [Official source / documentation](https://doi.org/10.1038/sdata.2014.22) Use the paper’s author-deposited Figshare data; document excluded/uncharacterized molecules, target property and units. | Follow the deposited dataset’s license and citation requirements; verify the exact release rather than a loader’s code license. |
| ZINC benchmark subsets | Molecular-graph regression; benchmark subsets and target definitions vary. Evaluation: State subset size, target and split; MAE for one benchmark configuration is not comparable to another target. | [Official source / documentation](https://github.com/graphdeeplearning/benchmarking-gnns) Use the benchmark authors’ versioned ZINC preparation and supplied splits; distinguish it from the entire ZINC chemical database. | Follow underlying ZINC terms and the chosen benchmark redistribution terms; model-code licensing is separate. |
| MUTAG | Binary graph classification on a small molecular dataset. Evaluation: Use nested/held-out model selection with repeated folds; do not select hyperparameters on the reported test folds. | [Official source / documentation](https://chrsmrrs.github.io/datasets/docs/datasets/) Use the TU collection’s MUTAG files and original-source bibliography; preserve graph labels and attributes. | Consult the collection and original source notices; no common blanket license is asserted for every TU dataset. |

Guide updated 2026-11-05.

## Key findings

- From Nov 2023 to Oct 2026, 102 archived papers mentioned Molecule benchmarks (QM9, ZINC, MUTAG) in their title, abstract, keywords or full text, most of them in Jan 2025 (6).
- No paper of the tracked journals mentioned Molecule benchmarks (QM9, ZINC, MUTAG) in the last 30 days (2026-10-07 to 2026-11-05).
- From 2017 to 2026 (so far), Molecule benchmarks (QM9, ZINC, MUTAG) was mentioned in 108 papers in the tracked journals, most of them in AI (105), information systems (2) and higher education and research (1).
- 31 of them were published in 2025, against 0 in 2017.
- Since 2023, the institutions with the most of these papers were Hong Kong Polytechnic University (6), Beihang University (4) and Shanghai Jiao Tong University (4).
- The journals publishing most of them were IEEE Transactions on Pattern Analysis and Machine Intelligence (15), IEEE Transactions on Knowledge and Data Engineering (13) and Neural Networks (12).
- The most-cited of them is “Augmenting large language models with chemistry tools” (Nature Machine Intelligence, 2024), with 795 citations.
- Papers use Molecule benchmarks (QM9, ZINC, MUTAG) most often with IMDb (28), Open Graph Benchmark (25) and Reddit (20); it has 15 partners in the code and data network, in the group around Twitter / X, Facebook, YouTube / TikTok.

## Mentions over 36 months

*Papers mentioning it by publication month, Nov 2023 to Oct 2026 (titles, abstracts, keywords and full texts)*

| Trend | Nov 23–Oct 24 | Nov 24–Oct 25 | Nov 25–Oct 26 | 36 months |
|---|---:|---:|---:|---:|
|  | 37 | 35 | 30 | 102 |

## Papers per year

*Papers per publication year and category*

| Category | 2017 | 2018 | 2019 | 2020 | 2021 | 2022 | 2023 | 2024 | 2025 | 2026 (so far) |
|---|---:|---:|---:|---:|---:|---:|---:|---:|---:|---:|
| [AI](https://scholar4ai.com/categories/ai.md) | 0 | 0 | 0 | 0 | 0 | 0 | 13 | 36 | 29 | 27 |
| [Information Systems](https://scholar4ai.com/categories/information-systems.md) | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 1 |
| [Higher Education and Research](https://scholar4ai.com/categories/higher-education-and-research.md) | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 |
| All categories | 0 | 0 | 0 | 0 | 0 | 0 | 13 | 36 | 31 | 28 |

## Where the papers come from

*Journals with the most papers, since 2023*

| # | Journal | Papers |
|---:|---|---:|
| 1 | [IEEE Transactions on Pattern Analysis and Machine Intelligence](https://scholar4ai.com/journals/ieee-transactions-on-pattern-analysis-and-machine-intelligence.md) | 15 |
| 2 | [IEEE Transactions on Knowledge and Data Engineering](https://scholar4ai.com/journals/ieee-transactions-on-knowledge-and-data-engineering.md) | 13 |
| 3 | [Neural Networks](https://scholar4ai.com/journals/neural-networks.md) | 12 |
| 4 | [Nature Machine Intelligence](https://scholar4ai.com/journals/nature-machine-intelligence.md) | 10 |
| 5 | [IEEE Transactions on Neural Networks and Learning Systems](https://scholar4ai.com/journals/ieee-transactions-on-neural-networks-and-learning-systems.md) | 8 |
| 6 | [Machine Learning](https://scholar4ai.com/journals/machine-learning.md) | 6 |
| 7 | [Pattern Recognition](https://scholar4ai.com/journals/pattern-recognition.md) | 5 |
| 8 | [Pattern Recognition Letters](https://scholar4ai.com/journals/pattern-recognition-letters.md) | 4 |
| 9 | [Neural Computing and Applications](https://scholar4ai.com/journals/neural-computing-and-applications.md) | 4 |
| 10 | [IEEE Transactions on Artificial Intelligence](https://scholar4ai.com/journals/ieee-transactions-on-artificial-intelligence.md) | 4 |

*Institutions with the most papers, since 2023*

| # | Institution | Papers |
|---:|---|---:|
| 1 | [Hong Kong Polytechnic University](https://scholar4ai.com/institutions/hong-kong-polytechnic-university.md) | 6 |
| 2 | [Beihang University](https://scholar4ai.com/institutions/beihang-university.md) | 4 |
| 3 | [Shanghai Jiao Tong University](https://scholar4ai.com/institutions/shanghai-jiao-tong-university.md) | 4 |
| 4 | [University of Science and Technology of China](https://scholar4ai.com/institutions/university-of-science-and-technology-of-china.md) | 4 |
| 5 | [Chinese Academy of Sciences](https://scholar4ai.com/institutions/chinese-academy-of-sciences.md) | 3 |
| 6 | [Chinese University of Hong Kong](https://scholar4ai.com/institutions/chinese-university-of-hong-kong.md) | 3 |
| 7 | [Harbin Institute of Technology](https://scholar4ai.com/institutions/harbin-institute-of-technology.md) | 3 |
| 8 | [Jilin University](https://scholar4ai.com/institutions/jilin-university.md) | 3 |
| 9 | [Macquarie University](https://scholar4ai.com/institutions/macquarie-university.md) | 3 |
| 10 | [Nanyang Technological University](https://scholar4ai.com/institutions/nanyang-technological-university.md) | 3 |

*Countries with the most papers, since 2023*

| # | Country | Papers |
|---:|---|---:|
| 1 | China | 57 |
| 2 | United States | 24 |
| 3 | Hong Kong | 14 |
| 4 | United Kingdom | 13 |
| 5 | Italy | 11 |
| 6 | Australia | 8 |
| 7 | Singapore | 7 |
| 8 | Switzerland | 5 |
| 9 | France | 4 |
| 10 | South Korea | 4 |

*Keywords with the most papers, since 2023*

| # | Keyword | Papers |
|---:|---|---:|
| 1 | [graph neural networks](https://scholar4ai.com/keywords/graph-neural-networks.md) | 55 |
| 2 | graph classification | 24 |
| 3 | graph representation learning | 13 |
| 4 | node classification | 11 |
| 5 | [interpretability](https://scholar4ai.com/keywords/interpretability.md) | 10 |
| 6 | molecular property prediction | 10 |
| 7 | over-smoothing | 10 |
| 8 | drug discovery | 8 |
| 9 | message passing | 8 |
| 10 | [graph convolutional networks](https://scholar4ai.com/keywords/graph-convolutional-networks.md) | 7 |

## Most-cited papers

*Most-cited papers since 2017*

| # | Citations | Per month | Title | Journal | Published | Authors |
|---:|---:|---:|---|---|---|---|
| 1 | 795 | 26.6 | [Augmenting large language models with chemistry tools](https://doi.org/10.1038/s42256-024-00832-8) | [Nature Machine Intelligence](https://scholar4ai.com/journals/nature-machine-intelligence.md) | 2024-05-08 | Andres M Bran, Sam Cox, Oliver Tobias Schilter et al. |
| 2 | 345 | 10.7 | [A Comprehensive Survey on Deep Graph Representation Learning](https://doi.org/10.1016/j.neunet.2024.106207) | [Neural Networks](https://scholar4ai.com/journals/neural-networks.md) | 2024-02-27 | Wei Ju, Zheng Fang, Yiyang Gu et al. |
| 3 | 183 | 5.3 | [Multi-modal molecule structure–text model for text-based retrieval and editing](https://doi.org/10.1038/s42256-023-00759-6) | [Nature Machine Intelligence](https://scholar4ai.com/journals/nature-machine-intelligence.md) | 2023-12-18 | Shengchao Liu, Weili Nie, Chengpeng Wang et al. |
| 4 | 161 | 9.8 | [A framework to evaluate machine learning crystal stability predictions](https://doi.org/10.1038/s42256-025-01055-1) | [Nature Machine Intelligence](https://scholar4ai.com/journals/nature-machine-intelligence.md) | 2025-06-23 | Janosh Riebesell, Rhys E. A. Goodall, Philipp Benner et al. |
| 5 | 125 | 3.4 | [Neural scaling of deep chemical models](https://doi.org/10.1038/s42256-023-00740-3) | [Nature Machine Intelligence](https://scholar4ai.com/journals/nature-machine-intelligence.md) | 2023-10-23 | Nathan C. Frey, Ryan Soklaski, Simon Axelrod et al. |

## Used together with

Papers use Molecule benchmarks (QM9, ZINC, MUTAG) most often with IMDb (28), Open Graph Benchmark (25) and Reddit (20); it has 15 partners in the code and data network, in the group around Twitter / X, Facebook, YouTube / TikTok.

*What papers use Molecule benchmarks (QM9, ZINC, MUTAG) together with (at least 3 papers; lift 1 = chance)*

| Used with | Type | Papers using both | Lift |
|---|---|---:|---:|
| [IMDb](https://scholar4ai.com/datasets/imdb.md) | Dataset | 28 | 52.8 |
| [Open Graph Benchmark](https://scholar4ai.com/datasets/open-graph-benchmark.md) | Dataset | 25 | 120.4 |
| [Reddit](https://scholar4ai.com/datasets/reddit.md) | Dataset | 20 | 14.6 |
| [MNIST / Fashion-MNIST](https://scholar4ai.com/datasets/mnist-fashion-mnist.md) | Dataset | 11 | 4.4 |
| [Cora / CiteSeer / PubMed (Planetoid)](https://scholar4ai.com/datasets/cora-citeseer-pubmed-planetoid.md) | Dataset | 9 | 35.3 |
| [ImageNet](https://scholar4ai.com/datasets/imagenet.md) | Dataset | 9 | 1.7 |
| [CIFAR-10 / CIFAR-100](https://scholar4ai.com/datasets/cifar-10-cifar-100.md) | Dataset | 7 | 2.2 |
| [Facebook](https://scholar4ai.com/datasets/facebook.md) | Dataset | 5 | 0.7 |
| [Twitter / X](https://scholar4ai.com/datasets/twitter-x.md) | Dataset | 4 | 0.4 |
| [Wikipedia](https://scholar4ai.com/datasets/wikipedia.md) | Dataset | 4 | 1.7 |
| [Amazon reviews](https://scholar4ai.com/datasets/amazon-reviews.md) | Dataset | 3 | 5.6 |
| [Hugging Face Hub](https://scholar4ai.com/datasets/hugging-face-hub.md) | Dataset | 3 | 2.9 |

The whole network: https://scholar4ai.com/code/network.md

## Cite this page

Scholar4AI. (2026). Molecule benchmarks (QM9, ZINC, MUTAG). Journal and Conference Research Rankings. Retrieved November 5, 2026, from https://scholar4ai.com/datasets/molecule-benchmarks-qm9-zinc-mutag

```bibtex
@misc{scholar4ai-datasets-molecule-benchmarks-qm9-zinc-mutag,
  author = {{Scholar4AI}},
  title = {Molecule benchmarks (QM9, ZINC, MUTAG)},
  year = {2026},
  howpublished = {\url{https://scholar4ai.com/datasets/molecule-benchmarks-qm9-zinc-mutag}},
  urldate = {2026-11-05},
  note = {Journal and Conference Research Rankings; data from OpenAlex (CC0); licensed CC BY-NC 4.0}
}
```

License: CC BY-NC 4.0 (https://creativecommons.org/licenses/by-nc/4.0/).