ArticlePloS one2019
Hidden bias in the DUD-E dataset leads to misleading performance of deep learning in structure-based virtual screening.
Article in PloS one, 2019. The graph could read no effect estimate from its abstract, so it casts no vote on the map. Cited by 102 papers.
What it found
Each row is one number read from the abstract, on the scale the paper reported it, with its interval. Left of the dashed line favours the treatment, right favours the comparator. Under each row is the sentence it came from. New to these charts? A ten-minute tutorial.
The abstract states no effect estimate the extractor could read, or names no intervention and outcome on the map, so this paper lights no cell and moves no belief. It is still indexed, cited and linked below.
The trial behind it
Trials whose registry record cites this paper, or whose number appears in the abstract. A trial that started after this paper was published is citing it as background, not reporting it.
Neither the registry nor the abstract names a trial number. If this is a trial report, that itself is worth knowing.
Who cites it
102 citing papers in PubMed.
- Evaluating Machine Learning Affinity Prediction Methods on Novel Lead Optimization Data: A 15-PGDH Case Study.ChemMedChem · 2026Article
- Physical priors improve performance of structure-based binding affinity models.bioRxiv : the preprint server for biology · 2026Article
- A Unified Hierarchical Multiscale Fusion Framework for Drug-Target Affinity Prediction: From Benchmark Performance to Nanomolar Inhibitor Discovery.Advanced science (Weinheim, Baden-Wurttemberg, Germany) · 2026Article
- How Advanced Artificial Intelligence Technologies Shape Drug-Drug and Drug-Target Interaction Modeling.Advanced science (Weinheim, Baden-Wurttemberg, Germany) · 2026Review
- Facilitating structure-based drug discovery with an artificial intelligence-driven virtual screening platform.Nature protocols · 2026Review
- ModBindProceedings of the National Academy of Sciences of the United States of America · 2026Article
- The Last Mile Problem: A Critical Assessment of Physics-Based and AI Tools for Small Molecule Binding Prediction in Virtual Screening.Journal of chemical information and modeling · 2026Article
- Enabling Automatic Generation of Protein-Ligand Complex Data Sets with Atomistic Detail.Journal of chemical information and modeling · 2026Article
- Integrative Computational Chemistry Approaches in Modern Drug Discovery: Advances in Docking, Pharmacophore Modeling, Molecular Dynamics, and Virtual Screening.Pharmaceutics · 2026Review
- Discussions on the generalization of HybridSP on more equivalent benchmarks.Briefings in bioinformatics · 2026Article
- Benchmark competitiveness and generalization are not equivalent: a critical commentary on HybridSP.Briefings in bioinformatics · 2026Article
- Could statistical potential models achieve comparable or better performance than deep learning models?Briefings in bioinformatics · 2026Article
- Article
- Machine Learning Application in Enhancing HIV Management and Treatment Outcomes: Revolutionizing HIV Infection.Current HIV research · 2026Review
- A Systematic Review of Drug-Related InteractionsUtilizing Deep Learning and LLMs for Prediction and Mitigation.ACS omega · 2025Review
- Identification of mathematical patterns in genomic spectrograms linked to variant classification in complete SARS-CoV-2 sequences.Scientific reports · 2025Article
- Predicting drug solubility in binary solvent mixtures using graph convolutional networks: a comprehensive deep learning approach.Scientific reports · 2025Article
- SCORCH2: A Generalized Heterogeneous Consensus Model for High-Enrichment Interaction-Based Virtual Screening.Advanced science (Weinheim, Baden-Wurttemberg, Germany) · 2025Article
- All That Glitters Is Not Gold: Importance of Rigorous Evaluation of Proteochemometric Models.Journal of chemical information and modeling · 2025Article
- Structure based drug design and machine learning approaches for identifying natural inhibitors against the human αβIII tubulin isotype.Scientific reports · 2025Article
42 more citing papers are in PubMed but not listed here.
Corrections and comments
PubMed lists nothing against this paper. Absence here is not a guarantee, only a check that was made.
Authors and funding
8 authors.
Funding
Abstract
Recently much effort has been invested in using convolutional neural network (CNN) models trained on 3D structural images of protein-ligand complexes to distinguish binding from non-binding ligands for virtual screening. However, the dearth of reliable protein-ligand x-ray structures and binding affinity data has required the use of constructed datasets for the training and evaluation of CNN molecular recognition models. Here, we outline various sources of bias in one such widely-used dataset, the Directory of Useful Decoys: Enhanced (DUD-E). We have constructed and performed tests to investigate whether CNN models developed using DUD-E are properly learning the underlying physics of molecular recognition, as intended, or are instead learning biases inherent in the dataset itself. We find that superior enrichment efficiency in CNN models can be attributed to the analogue and decoy bias hidden in the DUD-E dataset rather than successful generalization of the pattern of protein-ligand interactions. Comparing additional deep learning models trained on PDBbind datasets, we found that their enrichment performances using DUD-E are not superior to the performance of the docking program AutoDock Vina. Together, these results suggest that biases that could be present in constructed datasets should be thoroughly evaluated before applying them to machine learning based methodology development.
Indexed as
Identifiers
What OpenQuestion holds
Registered trials
Read under generation 80e0d062 · epoch 390. Bibliography from PubMed, PubMed Central and OpenAlex; grants from NIH RePORTER; trial links from ClinicalTrials.gov; estimates, votes and beliefs from the OpenQuestion graph.