Machine Learning-Driven Drug Discovery for TNBC: Dual IDO1 and TDO Inhibitors

A study by P. Priyanga, last updated in March 2026, employed a machine learning pipeline to identify dual inhibitors for cancer-related enzymes. The work is based on IC50 values from 1,037 distinct dual inhibitors sourced from ChEMBL and BindingDB databases. It resulted in the identification of two promising candidate compounds, NP000319 and NP003833, for potential therapeutic development.

Use Cases

• Training ML classification models for inhibitor activity based on IC50 values from public databases.
• Virtual screening of chemical libraries like MEGxp for potential dual IDO1/TDO inhibitors.
• Evaluating drug candidates using molecular docking, MM-GBSA calculations, and ADMET profiling.
• Assessing protein-ligand complex stability via molecular dynamics simulations data.

Strengths

• ML model reported 95% accuracy for classification.
• Training data includes IC50 values for 1,037 distinct dual inhibitors.
• Results include two identified lead compounds with predicted anticancer potential.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

figshare, author P. Priyanga.

Collection Method

Data sourced from ChEMBL and BindingDB databases, processed through an ML-driven virtual screening pipeline.

Freshness

Last updated 2026-03-17 15:20:11; freshness should be verified.