MolPILE -- large-scale, diverse dataset for molecular representation learning
Not on the map. Why?
Abstract (arXiv)
The size, diversity, and quality of pretraining datasets critically determine the generalization ability of foundation models. Despite their growing importance in chemoinformatics, the effectiveness of molecular representation learning has been hindered by limitations in existing small molecule datasets. To address this gap, we present MolPILE, large-scale, diverse, and rigorously curated collection of 222 million compounds, constructed from 6 large-scale databases using an automated curation pipeline. We present a comprehensive analysis of current pretraining datasets, highlighting considerable shortcomings for training ML models, and demonstrate how retraining existing models on MolPILE yields improvements in generalization performance. This work provides a standardized resource for model training, addressing the pressing need for an ImageNet-like dataset in molecular chemistry.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 31 of its 112 references that are in the map. First posted on arXiv 22 Sep 2025.
Builds on
31 of 112 references.
References · 31
- 2020Scaling Laws for Neural Language Modelscited by 6,586
- 2020The Pile: An 800GB Dataset of Diverse Text for Language Modelingcited by 2,411
- 2017MoleculeNet: A Benchmark for Molecular Machine Learningcited by 1,051
- 2019Strategies for Pre-training Graph Neural Networkscited by 971
- 2024Phi-4 Technical Reportcited by 711
- 2023DataComp: In search of the next generation of multimodal datasetscited by 644
- 2021Datasets: A Community Library for Natural Language Processingcited by 530
- 2021Therapeutics Data Commons: Machine Learning Datasets and Tasks for Drug Discovery and Developmentcited by 264
- 2021Pre-training Molecular Graph Representation with 3D Geometrycited by 248
- 2021Large-Scale Chemical Language Representations Capture Molecular Structure and Propertiescited by 206
- 2021ChemRL-GEM: Geometry Enhanced Molecular Representation Learning for Property Predictioncited by 204
- 2022ChemBERTa-2: Towards Chemical Foundation Modelscited by 127
19 more
- 2021Scaling Laws for Neural Machine Translationcited by 111
- 2020Molecular representation learning with language models and domain-relevant auxiliary taskscited by 94
- 2020Molecule Attention Transformercited by 92
- 2022SPICE, A Dataset of Drug-like Molecules and Peptides for Training Machine Learning Potentialscited by 82
- 2023Enhancing Activity Prediction Models in Drug Discovery with the Ability to Understand Human Languagecited by 44
- 2023SELFormer: Molecular Representation Learning via SELFIES Language Modelscited by 40
- 2021Chemical-Reaction-Aware Molecule Representation Learningcited by 36
- 2021How Much Space Has Been Explored? Measuring the Chemical Space Covered by Databases and Machine-Generated Moleculescited by 27
- 2021Relative Molecule Self-Attention Transformercited by 21
- 2025Benchmarking Pretrained Molecular Embedding Models For Molecular Representation Learningcited by 15
- 2024ChemFM as a Scaling Law Guided Foundation Model Pre-trained on Informative Chemicalscited by 10
- 2024Scikit-fingerprints: easy and efficient computation of molecular fingerprints in Pythoncited by 9
- 2024Transformers for molecular property prediction: Lessons learned from the past five yearscited by 8
- 2024ApisTox: a new benchmark dataset for the classification of small molecules toxicity on honey beescited by 6
- 2024Measuring Diversity: Axioms and Challengescited by 6
- 2024WelQrate: Defining the Gold Standard in Small Molecule Drug Discovery Benchmarkingcited by 5
- 2025Evaluating machine learning models for predicting pesticide toxicity to honey beescited by 2
- 2023Rapidgzip: Parallel Decompression and Seeking in Gzip Files Using Cache Prefetchingcited by 1
- 2025ToxBench: A Binding Affinity Prediction Benchmark with AB-FEP-Calculated Labels for Human Estrogen Receptor Alphacited by 1
Built on by
Citing papers per year: 2026: 1.
1 paper.