Sep 2026· Journal of Chemical Information and Modeling· Vol 66 18, pp.
11756-11766
· 0 citations· 43 references
Medicine
TL;DR
This work performs a systematic evaluation of graph neural networks (GNNs) using a ChEMBL-derived dataset spanning 5,368 targets and 1.59 million activity records, capturing the long-tailed distributions and target-specific imbalances commonly observed in pharmaceutical data.
Abstract
The performance of target-specific, ligand-based virtual screening models is strongly influenced by dataset characteristics, including data availability, class imbalance, and evaluation strategies. In this work, we perform a systematic evaluation of graph neural networks (GNNs) using a ChEMBL-derived dataset spanning 5,368 targets and 1.59 million activity records, capturing the long-tailed distributions and target-specific imbalances commonly observed in pharmaceutical data. Through a systematic evaluation of multiple GNN architectures, we identify guidelines for model selection: while the Graph Isomorphism Network (GIN) consistently outperforms others on datasets with >100 samples (a mean ROC-AUC up to 0.94), simpler architectures are more robust under extreme data scarcity. Critically, our comparative analysis of splitting strategies reveals that random sampling yields artificially optimistic performance due to structural overlaps, whereas similarity-aware clustering exposes a substantial generalization gap (AUC drop > 0.3), cautioning against prevailing evaluation practices. We further demonstrate that multi-task learning serves as an effective remedy for small-target instability, providing significant and consistent performance gains. To underscore its translational value, we deploy this comprehensive framework in a virtual screening campaign against Mcl-1, yielding a chemically optimized lead, C4 (Ki = 0.58 μM), with verified cellular efficacy. Our findings highlight the importance of task-aware benchmark design and offer a practical strategy for reliable GNN application in drug discovery.
A unified and standardized benchmarking framework for GDL-based toxicity prediction is introduced, systematically evaluating more than 20 representative approaches under consistent experimental conditions and across multiple datasets and partitioning strategies, enabling a fair and reproducible comparison of model perf...
Noel Suarez-Barro, M. Lama, J. C. Vidal· 0 citations
Hold-out evaluation, repeated seeds, group-aware bootstrap analysis, and a size-matched control support the same conclusion: architecture selection depends on the form of extrapolation, not only on average error or training-set size.
Minjae Chung, Clara Li, Malar Paavai Muthukumaran et al.· 0 citations
Molecular property prediction is a key task in AI-driven drug discovery, yet the prevalence and impact of label imbalances in molecular property regression remain poorly understood. Through a systematic benchmark of widely used molecular property data sets, we show that target values are often highly imbalanced and tha...
Y. Sun, Yu Shi, Alana Deng et al.· Journal of Chemical Informat...· 0 citations
This study proposes a new approach termed the Multi-Strategy Message Passing Graph Neural Network (MS-MPGN), which combines two complementary strategies: one that maps the overall shape of a molecule and another that focuses on local chemical interactions.
Ali Abdelkrim, Abdelkrim Bouramoul, Imene Zenbout· Iran Journal of Computer Sci...· 0 citations
This study investigates graph convolutional network prediction of protein-ligand binding affinity within the context of computational chemistry and machine learning for drug discovery, an area of growing scientific importance given its implications for virtual screening pipelines and lead optimization in pharmaceutical...
Aisha N. Patel, Etienne M. Beauchamp, Lucas H. Rodrigues· Princeton Journal of Pre-Col...· 0 citations
KAN-based post-aggregation transformations represent an emerging design option for graph-based DTA prediction, however, the latent-space and structure-based findings should not be interpreted as evidence of mechanistic biological relevance without prospective experimental validation.
Abla Bedoui, Nithyadevi Duraisamy, Mohammed Cherkaoui· Pharmaceuticals· 0 citations
Machine-learning networks can learn without detailed instructions from a designer. Those that do may offer insight into how the brain reorganizes itself during learning.
MIT News · Artificial Intelligence· news.mit.eduJul 15, 2026
Assistant Professor Pat Pataranutaporn describes a new interface that lets everyday users glimpse inside an AI's neural network before their chatbot ever says a word.
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.