Research Appraisalother

Machine Learning Approach to Anticancer Activity Prediction of Transition-Metal Complexes Based on a Large-Scale Experimental Database

Journal of medicinal chemistryKrasnov, Lev, Malikov, Dmitry, Kiseleva, Marina A et al.23 Apr 2026DOI

Clinical Snapshot

80CEBM

Evidence: Moderateother

PICO Framework

P — Population	Transition-metal complexes (ruthenium and iridium) tested against 754 cancer cell lines
I — Intervention	Machine learning-based cytotoxicity prediction models using metal-ligand composition data
C — Comparator	Experimental IC50 values from curated database (MetalCytoToxDB)
O — Outcomes	Cytotoxicity class prediction accuracy, external validation performance, model interpretability

Bottom Line

This computational study presents MetalCytoToxDB, a curated database of 26,500 IC50 values for metal complexes, and demonstrates machine learning models can predict cytotoxicity based solely on metal-ligand composition. While the approach shows promise for high-throughput screening of ruthenium and iridium complexes, the abstract lacks specific performance metrics. The publicly available database represents a valuable resource for medicinal chemists developing metal-based anticancer agents. However, clinicians should note this is purely computational work requiring experimental validation before clinical application. The methodology could accelerate early-stage drug discovery by reducing the need for extensive laboratory screening, though limitations in handling experimental variability and clinical correlation remain unclear from the available information.

Evidence: Moderate

Key Findings

P Value: Not reported
Effect Size: Not specified in abstract
Primary Outcome: Accurate cytotoxicity class prediction for ruthenium and iridium complexes
Nnt Or Sensitivity: External validation performed but specific metrics not provided
Confidence Interval: Not reported

Clinical Application

High - computational tool accessible online, requires minimal resources Relevant for Australian cancer research institutions and pharmaceutical companies developing novel metal-based therapeutics; aligns with TGA requirements for systematic drug development approaches Researchers developing metal-based anticancer agents, particularly ruthenium and iridium complexes

Abstract

In this work, we developed a straightforward data-driven approach to predict the cytotoxicity of metal complexes based entirely on their (metal + ligands) composition. To this end, we have manually curated MetalCytoToxDB─a comprehensive experimental database comprising 26,500 IC50 values for 7050 metal complexes against 754 cell lines from 1921 articles. Based on these, machine learning models were created to accurately assign the cytotoxicity class within the ruthenium and iridium subsets. Moreover, external validation of the best-performing model on the unseen data was carried out. The possibility of multimetal predictions was explored, enabling assessment of cytotoxicity among the complexes of metals, for which experimental data are relatively scarce. The interpretability and limitations of the developed models are discussed. Finally, a pipeline for the effective high-throughput computational screening of ruthenium complexes is proposed. The MetalCytoToxDB is available online for AI-assisted exploration at https://biometaldb.streamlit.app/.

References

1.Krasnov, L., Malikov, D., Kiseleva, M. A., Nykhrikova, E. V., Tatarin, S. V., & Bezzubov, S. I. (2026). Machine Learning Approach to Anticancer Activity Prediction of Transition-Metal Complexes Based on a Large-Scale Experimental Database. Journal of Medicinal Chemistry. https://doi.org/10.1021/acs.jmedchem.5c02755

Share:X LinkedIn

Related Research

Systematic Review

15CEBM

Cancer medicine

Applications of Artificial Intelligence in Cancer Diagnosis and Treatment

1 Aug 2026

other

65CEBM

Acta oncologica (Stockholm, Sweden)

User perceptions of machine learning models as decision support for colorectal cancer multidisciplinary team conferences (AID-SIM-2): a qualitative simulation study.