Pith. sign in

REVIEW 6 major objections 5 minor 65 references

M$^{3}$-20M: A Large-Scale Multi-Modal Molecule Dataset for AI-driven Drug Design and Discovery

T0 review · 6 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper claims that a 20-million-molecule dataset combining SMILES, graphs, 3D structures, properties, and text improves AI molecule generation and property prediction.

desk verdict Useful large multimodal molecule resource, but the headline '20M unique molecules' claim and the 'five modalities per molecule' promise need verification before the scale claims are taken at face value. read the letter →

arxiv 2412.06847 v2 pith:3YVQNLN6 submitted 2024-12-08 q-bio.QM cs.AIcs.LG

Siyuan Guo , Lexuan Wang , Chang Jin , Jinxian Wang , Han Peng , Huayang Shi , Wengen Li , Jihong Guan
show 1 more author
Shuigeng Zhou
This is my paper · ORCID
classification q-bio.QMcs.AIcs.LG
keywords multi-modalmoleculedatasetdrugdesigngenerationmolecularpropertypredictionlargelanguagemodelsSMILES3Dstructureconstruction
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper introduces M3-20M, an open-access dataset of more than 20 million molecules in which every molecule is accompanied by a SMILES string, a 2D molecular graph, a 3D structure, physicochemical properties, and a textual description. The authors' central claim is that this multi-modal coverage at this scale meaningfully improves AI models for drug design: using M3-20M as demonstration examples or fine-tuning data makes large language models and specialized generative models produce more valid, more diverse, and more novel molecules, and improves accuracy on molecular property prediction. The dataset is assembled by pulling molecules from PubChem, ZINC, and QM9, computing or fetching the structural and property modalities, and generating missing text with GPT-3.5 under an expert-scored quality check. If the claim holds, the dataset gives the field a common large-scale resource for training text-aware molecular models.

What carries the argument

The central object is the M3-20M dataset itself and its construction pipeline. Molecules are collected from PubChem, ZINC, and QM9; RDKit converts them to 2D graphs and reads 3D coordinates from PubChem SDF files; 26 physicochemical properties are pulled from PubChem and templated into sentences; missing text is generated by GPT-3.5 using a CO-STAR-style expert-chemist prompt, and a human expert scoring protocol (accuracy, effectiveness, comprehensiveness, simplicity) is used to regenerate low-quality descriptions. What carries the argument is the pairing of text with structure: it gives large language models a shared format in which property names and values become readable sentences, and the experiments compare models given SMILES only versus SMILES plus 3D coordinates plus text.

What would settle it

Count unique molecules in the released dataset by canonical SMILES or InChI keys; if the union of PubChem, ZINC, and QM9 entries drops below 20 million, the scale claim weakens. Then sample, say, 1,000 entries and test whether each has loadable 2D and 3D structures, non-empty text, and property values consistent with the SMILES; any large failure rate would undercut the claim that every molecule has all five modalities.

Watch

Extended reading notes

Core claim

M3-20M is claimed to be the largest open-access multi-modal molecule dataset for AI-driven drug design, containing over 20 million molecules and 71 times the molecule count of the largest prior multi-modal dataset. The paper argues that the combination of 1D SMILES, 2D graphs, 3D coordinates, physicochemical properties, and text descriptions changes what models can learn: in experiments, few-shot prompting with multi-modal examples and LoRA fine-tuning with multi-modal data consistently beat single-modal baselines on molecule generation (validity, uniqueness, novelty, and MOSES benchmark metrics) and on property prediction for QM9, BACE, BBBP, ClinTox, HIV, and Tox21. The dataset also ships seven augmented '-MM' sub-datasets for downstream tasks and tools for 2D graph image generation and PubMed text crawling.

Load-bearing premise

The load-bearing premise is that the 20 million molecules are distinct and that each one actually has valid data in every modality; the paper does not describe removing molecules shared among PubChem, ZINC, and QM9 or verifying each entry's 3D coordinates and text, so the effective scale and completeness could be lower than the headline claim.

Editorial extensions

If this is right

  • Models trained or prompted with M3-20M should generate molecules with higher validity, uniqueness, and novelty than with single-modal datasets; the paper reports this across GPT-3.5, GPT-4, GLM4, Llama3-8b, MoFlow, and D2L-OMP.
  • The seven '-MM' sub-datasets provide multi-modal versions of standard benchmarks, so molecule generation and property prediction results become comparable across researchers without each group rebuilding the modality alignment.
  • Because each molecule has a PubChem CID and text, the dataset can support text-to-molecule and molecule-to-text tasks, including captioning, retrieval, and text-guided editing, not only the two tasks tested.
  • The dataset is released under GPLv3 with bi-weekly maintenance, so downstream work can build on a stable, updating resource.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • An implication the authors leave implicit: if the observed gains come mainly from the textual descriptions, then smaller curated multi-modal sets with high-quality text might capture most of the benefit, and the 20-million scale matters less than the text quality; a controlled ablation varying text quality while holding molecule count fixed would test this.
  • Because the GPT-3.5-generated descriptions cover only about 0.9 percent of the dataset, the reproducibility of the text modality depends on the prompt and scoring protocol; a natural extension would be to release the generation prompts and validator instructions alongside the data.
  • Because the dataset merges PubChem, ZINC, and QM9 without a described deduplication step, a conservative reading is that the unique-molecule count may be lower than 20 million; computing canonical SMILES or InChI keys over the released files would settle the exact scale.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

6 major / 5 minor

Summary. This manuscript presents M3-20M, a multi-modal molecule dataset claimed to contain over 20 million molecules with SMILES strings, 2D graphs, 3D structures, physicochemical properties, and textual descriptions, assembled from PubChem, ZINC, and QM9, with text augmented by GPT-3.5 and filtered by human expert scoring. The paper also constructs seven '-MM' downstream datasets (QM9-MM, MOSES-MM, BACE-MM, BBBP-MM, HIV-MM, ClinTox-MM, Tox21-MM) and reports experiments on molecule generation and property prediction using closed- and open-source LLMs and specialized generative models. The central claims are that M3-20M is 71 times larger than the largest existing multi-modal dataset (PubChemSTM) and that multi-modal training significantly boosts model performance over single-modal data.

Significance. If the scale and modality-completeness claims are verified, M3-20M would be a valuable open resource for molecular machine learning, and the associated '-MM' benchmark datasets would support reproducible comparison. The paper's strengths include the open-data commitment, the explicit expert-scoring protocol for generated text, the breadth of evaluated models (GPT-3.5, GPT-4, GLM4, Llama3-8b, MoFlow, D2L-OMP), and the attempt to provide multiple downstream benchmarks. However, the central claims currently rest on unverified counts and a leakage-control procedure that is not described in sufficient detail, so the contribution's practical value cannot yet be assessed.

major comments (6)
  1. [Methodology, Data Collection] The construction pipeline merges PubChem, ZINC, and QM9 but no deduplication step is described. PubChem and ZINC are known to overlap, and QM9 is largely contained in PubChem's chemical space; if the released 20,249,090 entries are record counts rather than unique molecules, the headline 'over 20 million molecules' and the '71 times larger than PubChemSTM' comparison are inflated. Please report the number of unique molecules (e.g., by canonical SMILES or InChIKey) after merging, give the overlap matrix among sources, and recompute all downstream counts if duplicates are found.
  2. [Methodology, Data Collection; Table 10] The claim that each molecule has all five modalities is not currently supported. For molecules 'not present in the PubChem database', the paper does not say how 3D coordinates or physicochemical properties are obtained; 3D structure generation is described only through PubChem SDF downloads, and Table 10 shows property counts below 20M (e.g., Molecular Weight 19,174,000). Please provide per-modality coverage statistics (SMILES, graph, 3D, properties, text) and specify the procedure for generating 3D coordinates for non-PubChem molecules.
  3. [Text Description Generation by GPT-3.5; Statistic Analyses and Results] The description counts are internally inconsistent. The text lists three sources (direct PubChem texts, property-template conversion, and GPT-3.5), but the reported total 20,249,090 equals exactly the sum of 19,175,245 template descriptions and 1,073,845 GPT-generated descriptions, leaving no room for the 360,133 direct PubChem texts mentioned in the statistics section. In addition, 1,073,845 is 5.30% of 20,249,090, not 0.934% as stated. Please reconcile these numbers and provide a per-source breakdown of descriptions.
  4. [Molecular Property Prediction Experiments] The leakage-control claim is not verifiable. The paper states that 'we mask all the target property-related information in textual descriptions during training and evaluation', but no masking procedure, list of masked tokens/fields, or code is provided. Since a large fraction of the text was generated by GPT-3.5 and GPT-3.5 is itself one of the evaluated models, the reported gains could be inflated if target properties appear in the text. Please describe the masking protocol precisely and release the masking script; ideally, evaluate on a split where no evaluated model generated the training descriptions.
  5. [Experiments, Tables 6-8] The conclusion that multi-modal data 'consistently outperform' single-modal data is not supported by the reported numbers. In Table 6, GPT-4's dipole-moment MAE is worse with SMILES+3D+text (1.27) than with SMILES only (1.17), and adding 3D coordinates alone frequently degrades performance (e.g., GLM-4 alpha: 12.82 to 14.53). In Table 8, 4 of 12 Tox21 subtasks show lower accuracy in the multi-modal setting (NR-AR-LBD, SR-ARE, SR-ATAD5, SR-p53). In Table 7, the BBBP-MM difference (0.228 vs 0.272) is accompanied by standard deviations of 0.02-0.03 and is far below typical supervised baselines for this benchmark. Please report statistical significance tests, discuss the mixed results explicitly, and moderate the wording of the central performance claim.
  6. [Multi-Modal Datasets for Downstream Tasks, Table 2] QM9-MM is listed with only 6,830 molecules although QM9 contains 134K molecules; no filtering criterion is given. Because the molecule-generation and regression experiments are run on QM9-MM, the reader cannot tell whether the results reflect the full QM9 chemical space or a small curated subset. Please explain how the 6,830 molecules were selected and whether the same selection affects the comparison to single-modal baselines.
minor comments (5)
  1. [Molecule Generation Experiments, Experimental Details] The paragraph beginning 'We employ the FP-ICL method...' is duplicated verbatim; please remove one copy.
  2. [Table 9 caption] The caption contains a typo: 'in terns of ACC mean' should be 'in terms of ACC mean'.
  3. [Tables 1 and 5 and related text] The dataset name 'PCdes' is written as both 'PCdes' and 'Pcdes'; please standardize the spelling.
  4. [Generated Text Quality Control by Human Expert Scoring] The sentence 'For every 100 generated molecules, 10% are randomly selected' is imprecise; it should state explicitly that 10% of generated descriptions were sampled for expert scoring.
  5. [Data and Software Availability Statement] The Data and Software Availability Statement does not contain the repository URL; please include the URL in that section rather than only in the abstract.

Circularity Check

0 steps flagged · score 1.0 of 10

No circular derivation; dataset construction and evaluations are self-contained, with only a non-load-bearing self-citation.

full rationale

M3-20M is a dataset construction paper; there is no mathematical derivation whose conclusion equals its assumptions. The five modalities are assembled from PubChem, ZINC, QM9, RDKit conversions, and GPT-3.5 text generation; the downstream benchmarks (MOSES, MoleculeNet-style tasks, QM9-MM) compare multi-modal versus single-modal inputs and do not fit any parameter to the claimed outcome. The property prediction protocol explicitly states that target-property-related information is masked, so even though masking is not code-verified, the paper does not define the prediction in terms of the generated texts. The only author-overlap citation is D2L-OMP [23] used as one of several baselines; its inclusion does not carry the central claim that the dataset is large or useful. Concerns about duplicate counts across PubChem/ZINC/QM9 and about unverified masking are data-quality and leakage risks, not circularity under the definitions used here.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The central claim rests on the reliability of external databases, the correctness of RDKit computations, the effectiveness of a sampling-based quality check, and an unverified masking step. No new fitted parameters or invented entities are introduced.

assumptions (4)
  • domain assumption Source databases (PubChem, ZINC, QM9) provide accurate molecular structures and properties.
    The dataset is built by integrating these databases without independent verification of each record, as described in Methodology, Data Collection.
  • domain assumption RDKit correctly computes 2D graphs and 3D coordinates from SMILES/SDF data.
    The paper uses RDKit functions (Chem, GetAtomPosition) for graph and coordinate extraction, assuming their correctness.
  • ad hoc to paper GPT-3.5-generated descriptions are scientifically accurate after the expert scoring filter.
    The quality control samples 12,000 of 1,073,845 generated descriptions and extrapolates the pass rate; this assumes the stratified sampling and the scoring protocol are representative.
  • domain assumption Masking target-property information in text descriptions prevents data leakage during property prediction.
    The paper states this masking is performed, but no verification or code is provided, so the assumption is unverified.

how reviews work

0 comments
Cite this review

Pith. "Pith review of M$^{3}$-20M: A Large-Scale Multi-Modal Molecule Dataset for AI-driven Drug Design and Discovery." pith.science (2026). https://pith.science/paper/3YVQNLN6

@misc{pith2026241206847,
  author       = {Pith},
  title        = {Pith review of: M$^3$-20M: A Large-Scale Multi-Modal Molecule Dataset for AI-driven Drug Design and Discovery},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3YVQNLN6}},
  note         = {Machine review of arXiv:2412.06847}
}
abstract

This paper introduces M$^{3}$-20M, a large-scale Multi-Modal Molecule dataset that contains over 20 million molecules, with the data mainly being integrated from existing databases and partially generated by large language models. Designed to support AI-driven drug design and discovery, M$^{3}$-20M is 71 times more in the number of molecules than the largest existing dataset, providing an unprecedented scale that can highly benefit the training or fine-tuning of models, including large language models for drug design and discovery tasks. This dataset integrates one-dimensional SMILES, two-dimensional molecular graphs, three-dimensional molecular structures, physicochemical properties, and textual descriptions collected through web crawling and generated using GPT-3.5, offering a comprehensive view of each molecule. To demonstrate the power of M$^{3}$-20M in drug design and discovery, we conduct extensive experiments on two key tasks: molecule generation and molecular property prediction, using large language models including GLM4, GPT-3.5, GPT-4, and Llama3-8b. Our experimental results show that M$^{3}$-20M can significantly boost model performance in both tasks. Specifically, it enables the models to generate more diverse and valid molecular structures and achieve higher property prediction accuracy than existing single-modal datasets, which validates the value and potential of M$^{3}$-20M in supporting AI-driven drug design and discovery. The dataset is available at https://github.com/bz99bz/M-3.

Figures

Figures reproduced from arXiv: 2412.06847 by the authors.

Figure 1
Figure 1. Some examples from our M3 -20M dataset [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. The construction pipeline of our multi-modal molecule dataset M3 -20M [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Molecular text description generation by GPT-3.5. including structure, physicochemical properties, reactivity, and potential applications. b. Fairly Comprehensive (1 point): Several aspects of the molecule are described, though some areas may lack depth. c. Incomplete (0 points): The description is limited to one aspect of the molecule and significantly lacks breadth. 4. Simplicity (Maximum 1 point) a. Concise (1 po… view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Paradigms of model training and finetuning using our dataset [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: shows the 20 most frequently occurring words in textual descriptions of our dataset, including “natural”, “product”, “acid”, and “metabolite” etc. These terms appear tens of thousands of times, reflecting the dataset’s emphasis on biologically derived compounds. The fr…
Figure 6
Figure 6. Figure 6: Visualization of molecular space with uniform manifold approximation and projection (UMAP). Conclusion In this paper, we introduce M3 -20M, a new and large￾scale multi-modal molecule dataset containing over 20 million molecules. Extensive experiments demonstrate M3 -20…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

65 extracted references · 47 canonical work pages

  1. [1]

    Gpt-4 technical report

    Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, et al. Gpt-4 technical report. arXiv preprint arXiv:2303.08774, 2023

    arXiv 2023

  2. [2]

    Extracting structured data from organic synthesis procedures using a fine-tuned large language model

    Qianxiang Ai, Fanwang Meng, Jiale Shi, Brenden Pelkie, and Connor W Coley. Extracting structured data from organic synthesis procedures using a fine-tuned large language model. 2024

    work page 2024

  3. [3]

    Geom, energy-annotated molecular conformations for property prediction and molecular generation

    Simon Axelrod and Rafael Gomez-Bombarelli. Geom, energy-annotated molecular conformations for property prediction and molecular generation. Scientific Data , 9(1):185, 2022

    work page 2022

  4. [4]

    New substructure filters for removal of pan assay interference compounds (pains) from screening libraries and for their exclusion in bioassays

    Jonathan B Baell and Georgina A Holloway. New substructure filters for removal of pan assay interference compounds (pains) from screening libraries and for their exclusion in bioassays. Journal of medicinal chemistry , 53(7):2719–2740, 2010

    work page 2010

  5. [5]

    Molgpt: molecular generation using a transformer-decoder model

    Viraj Bagal, Rishal Aggarwal, PK Vinod, and U Deva Priyakumar. Molgpt: molecular generation using a transformer-decoder model. Journal of Chemical Information and Modeling , 62(9):2064–2076, 2021

    work page 2021

  6. [6]

    Towards foundational models for molecular learning on large-scale multi-task datasets

    Dominique Beaini, Shenyang Huang, Joao Alex Cunha, Zhiyi Li, Gabriela Moisescu-Pareja, Oleksandr Dymov, Samuel Maddrell-Mander, Callum McLean, Frederik Wenkel, Luis M¨ uller, et al. Towards foundational models for molecular learning on large-scale multi-task datasets. arXiv preprint arXiv:2310.04292, 2023

    arXiv

  7. [7]

    The properties of known drugs

    Guy W Bemis and Mark A Murcko. The properties of known drugs. 1. molecular frameworks. Journal of medicinal chemistry, 39(15):2887–2893, 1996

    1996

  8. [8]

    Language models are few-shot learners

    Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, et al. Language models are few-shot learners. Advances in neural information processing systems , 33:1877–1901, 2020

    1901

Show all 65 references
  1. [9]

    Artificial intelligence for drug discovery: Resources, methods, and applications

    Wei Chen, Xuesong Liu, Sanyin Zhang, and Shilin Chen. Artificial intelligence for drug discovery: Resources, methods, and applications. Molecular Therapy-Nucleic Acids, 31:691–702, 2023

    2023

  2. [10]

    Chemberta: large-scale self-supervised pretraining for molecular property prediction

    Seyone Chithrananda, Gabriel Grand, and Bharath Ramsundar. Chemberta: large-scale self-supervised pretraining for molecular property prediction. arXiv preprint arXiv:2010.09885, 2020

    2010 arXiv

  3. [11]

    Molecular representations in ai-driven drug discovery: a review and practical guide

    Laurianne David, Amol Thakkar, Roc ´ ıo Mercado, and Ola Engkvist. Molecular representations in ai-driven drug discovery: a review and practical guide. Journal of Cheminformatics, 12(1):56, 2020

    2020

  4. [12]

    On the art of compiling and using’drug-like’chemical fragment spaces

    Jorg Degen, Christof Wegscheid-Gerlach, Andrea Zaliani, and Matthias Rarey. On the art of compiling and using’drug-like’chemical fragment spaces. ChemMedChem, 3(10):1503, 2008

    2008

  5. [13]

    Glm: General language model pretraining with autoregressive blank infilling

    Zhengxiao Du, Yujie Qian, Xiao Liu, Ming Ding, Jiezhong Qiu, Zhilin Yang, and Jie Tang. Glm: General language model pretraining with autoregressive blank infilling. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Paper...

    2022

  6. [14]

    The llama 3 herd of models

    Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Amy Yang, Angela Fan, et al. The llama 3 herd of models. arXiv preprint arXiv:2407.21783 , 2024. 12 Author Name et al

    2024 arXiv

  7. [15]

    Molecular structure input on the web

    Peter Ertl. Molecular structure input on the web. Journal of cheminformatics , 2:1–9, 2010

    2010

  8. [16]

    Molecular representation learning with language models and domain-relevant auxiliary tasks

    Benedek Fabian, Thomas Edlich, H´ el´ ena Gaspar, Marwin Segler, Joshua Meyers, Marco Fiscato, and Mohamed Ahmed. Molecular representation learning with language models and domain-relevant auxiliary tasks. arXiv preprint arXiv:2011.13230, 2020

    2011 arXiv

  9. [17]

    Tobias Fink and Jean-Louis Reymond. Virtual exploration of the chemical universe up to 11 atoms of c, n, o, f: assembly of 26.4 million structures (110.9 million stereoisomers) and analysis for new ring systems, stereochemistry, physicochemical properties, compound classes, an...

  10. [18]

    Probabilistic transformer: Modelling ambiguities and distributions for rna folding and molecule design

    J¨ org Franke, Frederic Runge, and Frank Hutter. Probabilistic transformer: Modelling ambiguities and distributions for rna folding and molecule design. Advances in Neural Information Processing Systems , 35:26856– 26873, 2022

    2022

  11. [19]

    Admetlab 3.0: an updated comprehensive online admet prediction platform enhanced with broader coverage, improved performance, api functionality and decision support

    Li Fu, Shaohua Shi, Jiacai Yi, Ningning Wang, Yuanhang He, Zhenxing Wu, Jinfu Peng, Youchao Deng, Wenxuan Wang, Chengkun Wu, et al. Admetlab 3.0: an updated comprehensive online admet prediction platform enhanced with broader coverage, improved performance, api functionality a...

    2024

  12. [20]

    Sample efficiency matters: a benchmark for practical molecular optimization

    Wenhao Gao, Tianfan Fu, Jimeng Sun, and Connor Coley. Sample efficiency matters: a benchmark for practical molecular optimization. Advances in neural information processing systems, 35:21342–21357, 2022

    2022

  13. [21]

    Chembl: a large-scale bioactivity database for drug discovery

    Anna Gaulton, Louisa J Bellis, A Patricia Bento, Jon Chambers, Mark Davies, Anne Hersey, Yvonne Light, Shaun McGlinchey, David Michalovich, Bissan Al-Lazikani, et al. Chembl: a large-scale bioactivity database for drug discovery. Nucleic acids research , 40(D1):D1100–D1107, 2012

    2012

  14. [22]

    Prefix-tree decoding for predicting mass spectra from molecules

    Samuel Goldman, John Bradshaw, Jiayi Xin, and Connor Coley. Prefix-tree decoding for predicting mass spectra from molecules. Advances in Neural Information Processing Systems, 36:48548–48572, 2023

    2023

  15. [23]

    Diffusing on two levels and optimizing for multiple properties: A novel approach to generating molecules with desirable properties

    Siyuan Guo, Jihong Guan, and Shuigeng Zhou. Diffusing on two levels and optimizing for multiple properties: A novel approach to generating molecules with desirable properties. IEEE/ACM Transactions on Computational Biology and Bioinformatics, 2024

    2024

  16. [24]

    What indeed can gpt models do in chemistry? a comprehensive benchmark on eight tasks

    Taicheng Guo, Kehan Guo, Zhengwen Liang, Zhichun Guo, Nitesh V Chawla, Olaf Wiest, Xiangliang Zhang, et al. What indeed can gpt models do in chemistry? a comprehensive benchmark on eight tasks. arXiv preprint arXiv:2305.18365, 16, 2023

    2023 arXiv

  17. [25]

    Shion Honda, Shoi Shi, and Hiroki R. Ueda. Smiles transformer: Pre-trained molecular fingerprint for low data drug discovery. 2019

    2019

  18. [26]

    Lora: Low-rank adaptation of large language models

    Edward J Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen- Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. Lora: Low-rank adaptation of large language models. arXiv preprint arXiv:2106.09685, 2021

    2021 arXiv

  19. [27]

    Principles of early drug discovery

    James P Hughes, Stephen Rees, S Barrett Kalindjian, and Karen L Philpott. Principles of early drug discovery. British journal of pharmacology , 162(6):1239–1249, 2011

    2011

  20. [28]

    Zinc- a free database of commercially available compounds for virtual screening

    John J Irwin and Brian K Shoichet. Zinc- a free database of commercially available compounds for virtual screening. Journal of chemical information and modeling, 45(1):177– 182, 2005

    2005

  21. [29]

    Comprehensive assessment of nine target prediction web services: which should we choose for target fishing? Briefings in Bioinformatics, 24(2):bbad014, 2023

    Kai-Yue Ji, Chong Liu, Zhao-Qian Liu, Ya-Feng Deng, Ting-Jun Hou, and Dong-Sheng Cao. Comprehensive assessment of nine target prediction web services: which should we choose for target fishing? Briefings in Bioinformatics, 24(2):bbad014, 2023

    2023

  22. [30]

    Pubchem substance and compound databases

    Sunghwan Kim, Paul A Thiessen, Evan E Bolton, Jie Chen, Gang Fu, Asta Gindulyte, Lianyi Han, Jane He, Siqian He, Benjamin A Shoemaker, et al. Pubchem substance and compound databases. Nucleic acids research, 44(D1):D1202–D1213, 2016

    2016

  23. [31]

    Molecule generation by principal subgraph mining and assembling

    Xiangzhe Kong, Wenbing Huang, Zhixing Tan, and Yang Liu. Molecule generation by principal subgraph mining and assembling. Advances in Neural Information Processing Systems, 35:2550–2563, 2022

    2022

  24. [32]

    Natural questions: a benchmark for question answering research

    Tom Kwiatkowski, Jennimaria Palomaki, Olivia Redfield, Michael Collins, Ankur Parikh, Chris Alberti, Danielle Epstein, Illia Polosukhin, Jacob Devlin, Kenton Lee, et al. Natural questions: a benchmark for question answering research. Transactions of the Association for Computa...

    2019

  25. [33]

    Rdkit documentation

    Greg Landrum. Rdkit documentation. Release, 1(1-79):4, 2013

    2013

  26. [34]

    Effective drug–target interaction prediction with mutual interaction neural network

    Fei Li, Ziqiao Zhang, Jihong Guan, and Shuigeng Zhou. Effective drug–target interaction prediction with mutual interaction neural network. Bioinformatics, 38(14):3582– 3589, 2022

    2022

  27. [35]

    Deep learning methods for molecular representation and property prediction

    Zhen Li, Mingjian Jiang, Shuang Wang, and Shugang Zhang. Deep learning methods for molecular representation and property prediction. Drug Discovery Today , 27(12):103373, 2022

    2022

  28. [36]

    Git- mol: A multi-modal large language model for molecular science with graph, image, and text

    Pengfei Liu, Yiming Ren, Jun Tao, and Zhixiang Ren. Git- mol: A multi-modal large language model for molecular science with graph, image, and text. Computers in Biology and Medicine, 171:108073, 2024

    2024

  29. [37]

    A quantitative analysis of knowledge-learning preferences in large language models in molecular science

    Pengfei Liu, Jun Tao, and Zhixiang Ren. A quantitative analysis of knowledge-learning preferences in large language models in molecular science. Nature Machine Intelligence , pages 1–13, 2025

    2025

  30. [38]

    A group symmetric stochastic differential equation model for molecule multi-modal pretraining

    Shengchao Liu, Weitao Du, Zhi-Ming Ma, Hongyu Guo, and Jian Tang. A group symmetric stochastic differential equation model for molecule multi-modal pretraining. In International Conference on Machine Learning , pages 21497–21526. PMLR, 2023

    2023

  31. [39]

    Multi-modal molecule structure–text model for text-based retrieval and editing

    Shengchao Liu, Weili Nie, Chengpeng Wang, Jiarui Lu, Zhuoran Qiao, Ling Liu, Jian Tang, Chaowei Xiao, and Animashree Anandkumar. Multi-modal molecule structure–text model for text-based retrieval and editing. Nature Machine Intelligence , 5(12):1447–1457, 2023

    2023

  32. [40]

    Dynamicbind: predicting ligand-specific protein-ligand complex structure with a deep equivariant generative model

    Wei Lu, Jixian Zhang, Weifeng Huang, Ziqiao Zhang, Xiangyu Jia, Zhenyu Wang, Leilei Shi, Chengtao Li, Peter G Wolynes, and Shuangjia Zheng. Dynamicbind: predicting ligand-specific protein-ligand complex structure with a deep equivariant generative model. Nature Communications,...

    2024

  33. [41]

    Gps++: An optimised hybrid mpnn/transformer for molecular property prediction

    Dominic Masters, Josef Dean, Kerstin Klaser, Zhiyi Li, Sam Maddrell-Mander, Adam Sanders, Hatem Helal, Deniz Beker, Ladislav Ramp´ aˇ sek, and Dominique Beaini. Gps++: An optimised hybrid mpnn/transformer for molecular property prediction. arXiv preprint arXiv:2212.02229 , 2022

    2022 arXiv

  34. [42]

    The natural language decathlon: Multitask learning as question answering

    Bryan McCann, Nitish Shirish Keskar, Caiming Xiong, and Richard Socher. The natural language decathlon: Multitask learning as question answering. arXiv preprint arXiv:1806.08730, 2018. Short Article Title 13

    2018 arXiv

  35. [43]

    Umap: Uniform manifold approximation and projection for dimension reduction

    Leland McInnes, John Healy, and James Melville. Umap: Uniform manifold approximation and projection for dimension reduction. arXiv preprint arXiv:1802.03426 , 2018

    2018 arXiv

  36. [44]

    Open babel: An open chemical toolbox

    Noel M O’Boyle, Michael Banck, Craig A James, Chris Morley, Tim Vandermeersch, and Geoffrey R Hutchison. Open babel: An open chemical toolbox. Journal of cheminformatics, 3:1–14, 2011

    2011

  37. [45]

    Training language models to follow instructions with human feedback

    Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, et al. Training language models to follow instructions with human feedback. Advances in neural information processing systems, 35:2773...

    2022

  38. [46]

    Molecular sets (moses): a benchmarking platform for molecular generation models

    Daniil Polykovskiy, Alexander Zhebrak, Benjamin Sanchez- Lengeling, Sergey Golovanov, Oktai Tatanov, Stanislav Belyaev, Rauf Kurbanov, Aleksey Artamonov, Vladimir Aladinskiy, Mark Veselov, et al. Molecular sets (moses): a benchmarking platform for molecular generation models. ...

    2020

  39. [47]

    A call for clarity in reporting bleu scores

    Matt Post. A call for clarity in reporting bleu scores. arXiv preprint arXiv:1804.08771, 2018

    2018 arXiv

  40. [48]

    Quantum chemistry structures and properties of 134 kilo molecules

    Raghunathan Ramakrishnan, Pavlo O Dral, Matthias Rupp, and O Anatole Von Lilienfeld. Quantum chemistry structures and properties of 134 kilo molecules. Scientific data, 1(1):1–7, 2014

    2014

  41. [49]

    Self-supervised graph transformer on large-scale molecular data

    Yu Rong, Yatao Bian, Tingyang Xu, Weiyang Xie, Ying Wei, Wenbing Huang, and Junzhou Huang. Self-supervised graph transformer on large-scale molecular data. Advances in neural information processing systems, 33:12559–12571, 2020

    2020

  42. [50]

    Enumeration of 166 billion organic small molecules in the chemical universe database gdb-17

    Lars Ruddigkeit, Ruud Van Deursen, Lorenz C Blum, and Jean-Louis Reymond. Enumeration of 166 billion organic small molecules in the chemical universe database gdb-17. Journal of chemical information and modeling , 52(11):2864–2875, 2012

    2012

  43. [51]

    A comprehensive map of molecular drug targets

    Rita Santos, Oleg Ursu, Anna Gaulton, A Patr ´ ıcia Bento, Ramesh S Donadi, Cristian G Bologa, Anneli Karlsson, Bissan Al-Lazikani, Anne Hersey, Tudor I Oprea, et al. A comprehensive map of molecular drug targets. Nature reviews Drug discovery , 16(1):19–34, 2017

    2017

  44. [52]

    Equivariant flow matching with hybrid probability transport for 3d molecule generation

    Yuxuan Song, Jingjing Gong, Minkai Xu, Ziyao Cao, Yanyan Lan, Stefano Ermon, Hao Zhou, and Wei-Ying Ma. Equivariant flow matching with hybrid probability transport for 3d molecule generation. Advances in Neural Information Processing Systems, 36, 2024

    2024

  45. [53]

    Evidence-based absorption, distribution, metabolism, excretion (adme) and its interplay with alternative toxicity methods

    Katya Tsaioun, Bas J Blaauboer, and Thomas Hartung. Evidence-based absorption, distribution, metabolism, excretion (adme) and its interplay with alternative toxicity methods. ALTEX-Alternatives to animal experimentation, 33(4):343–358, 2016

    2016

  46. [54]

    Glue: A multi- task benchmark and analysis platform for natural language understanding

    Alex Wang, Amanpreet Singh, Julian Michael, Felix Hill, Omer Levy, and Samuel R Bowman. Glue: A multi- task benchmark and analysis platform for natural language understanding. arXiv preprint arXiv:1804.07461 , 2018

    2018 arXiv

  47. [55]

    Deep learning approaches for de novo drug design: An overview

    Mingyang Wang, Zhe Wang, Huiyong Sun, Jike Wang, Chao Shen, Gaoqi Weng, Xin Chai, Honglin Li, Dongsheng Cao, and Tingjun Hou. Deep learning approaches for de novo drug design: An overview. Current opinion in structural biology, 72:135–144, 2022

    2022

  48. [56]

    Multitask joint strategies of self- supervised representation learning on biomedical networks for drug discovery

    Xiaoqi Wang, Yingjie Cheng, Yaning Yang, Yue Yu, Fei Li, and Shaoliang Peng. Multitask joint strategies of self- supervised representation learning on biomedical networks for drug discovery. Nature Machine Intelligence, 5(4):445– 456, 2023

    2023

  49. [57]

    Multimodal-learning for predicting molecular properties: A framework based on image and graph structures

    Zhuoyuan Wang, Jiacong Mi, Shan Lu, and Jieyue He. Multimodal-learning for predicting molecular properties: A framework based on image and graph structures. arXiv preprint arXiv:2311.16666, 2023

    2023 arXiv

  50. [58]

    Moleculenet: a benchmark for molecular machine learning

    Zhenqin Wu, Bharath Ramsundar, Evan N Feinberg, Joseph Gomes, Caleb Geniesse, Aneesh S Pappu, Karl Leswing, and Vijay Pande. Moleculenet: a benchmark for molecular machine learning. Chemical science, 9(2):513– 530, 2018

    2018

  51. [59]

    Hit and lead discovery with explorative rl and fragment-based molecule generation

    Soojung Yang, Doyeong Hwang, Seul Lee, Seongok Ryu, and Sung Ju Hwang. Hit and lead discovery with explorative rl and fragment-based molecule generation. Advances in Neural Information Processing Systems , 34:7924–7936, 2021

    2021

  52. [60]

    Quandb: a quantum chemical property database towards enhancing 3d molecular representation learning

    Zhijiang Yang, Tengxin Huang, Li Pan, Jingjing Wang, Liangliang Wang, Junjie Ding, and Junhua Xiao. Quandb: a quantum chemical property database towards enhancing 3d molecular representation learning. Journal of Cheminformatics, 16(1):48, 2024

    2024

  53. [61]

    Drugassist: A large language model for molecule optimization

    Geyan Ye, Xibao Cai, Houtim Lai, Xing Wang, Junhong Huang, Longyue Wang, Wei Liu, and Xiangxiang Zeng. Drugassist: A large language model for molecule optimization. Briefings in Bioinformatics , 26(1):bbae693, 2025

    2025

  54. [62]

    A unified drug–target interaction prediction framework based on knowledge graph and recommendation system

    Qing Ye, Chang-Yu Hsieh, Ziyi Yang, Yu Kang, Jiming Chen, Dongsheng Cao, Shibo He, and Tingjun Hou. A unified drug–target interaction prediction framework based on knowledge graph and recommendation system. Nature communications, 12(1):6775, 2021

    2021

  55. [63]

    Multimodal molecular pretraining via modality blending

    Qiying Yu, Yudi Zhang, Yuyan Ni, Shikun Feng, Yanyan Lan, Hao Zhou, and Jingjing Liu. Multimodal molecular pretraining via modality blending. In The Twelfth International Conference on Learning Representations , 2023

    2023

  56. [64]

    Moflow: an invertible flow model for generating molecular graphs

    Chengxi Zang and Fei Wang. Moflow: an invertible flow model for generating molecular graphs. In Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining , pages 617–626, 2020

    2020

  57. [65]

    A deep-learning system bridging molecule structure and biomedical text with comprehension comparable to human professionals

    Zheni Zeng, Yuan Yao, Zhiyuan Liu, and Maosong Sun. A deep-learning system bridging molecule structure and biomedical text with comprehension comparable to human professionals. Nature communications, 13(1):862, 2022

    2022

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.