Related Experiment Video
Updated: Jul 12, 2025
Network Pharmacology Prediction and Experimental Validation of Trichosanthes-Fritillaria thunbergii Action Mechanism Against Lung Adenocarcinoma
Published on: March 3, 2023
Classification models for predicting the bioactivity of pan-TRK inhibitors and SAR analysis
Xiaoman Zhao1,2, Yue Kong1, Yueshan Ji1
1College of Life Science and Technology, Beijing University of Chemical Technology, 15 BeiSanHuan East Road, Beijing, 100029, People's Republic of China.
Abstract:
Tropomyosin receptor kinases (TRKs) are important broad-spectrum anticancer targets. The oncogenic rearrangement of the NTRK gene disrupts the extracellular structural domain and epitopes for therapeutic antibodies, making small-molecule inhibitors essential for treating NTRK fusion-driven tumors. In this work, several algorithms were used to construct descriptor-based and nondescriptor-based models, and the models were evaluated by outer 10-fold cross-validation. To find a model with good generalization ability, the dataset was partitioned by random and cluster-splitting methods to construct in- and cross-domain models, respectively. Among the 48 models built, the model with the combination of the deep neural network (DNN) algorithm and extended connectivity fingerprints 4 (ECFP4) descriptors achieved excellent performance in both dataset divisions. The results indicate that the DNN algorithm has a strong generalization prediction ability, and the richness of features plays a vital role in predicting unknown spatial molecules. Additionally, we combined the clustering results and decision tree models of fingerprint descriptors to perform structure-activity relationship analysis. It was found that nitrogen-containing aromatic heterocyclic and benzo heterocyclic structures play a crucial role in enhancing the activity of TRK inhibitors.
Insights
Researchers developed computational models to identify new TRK inhibitors for NTRK fusion cancers. A deep neural network with ECFP4 descriptors showed strong predictive power, highlighting the importance of molecular features and heterocyclic structures for drug discovery.
Area of Science:
- Medicinal Chemistry
- Computational Drug Discovery
- Oncology
Background:
- Tropomyosin receptor kinases (TRKs) are key targets in treating NTRK fusion-driven cancers.
- NTRK gene rearrangements necessitate small-molecule inhibitors due to antibody epitope disruption.
- Developing effective TRK inhibitors requires robust predictive modeling.
Purpose of the Study:
- To construct and evaluate computational models for predicting TRK inhibitor activity.
- To identify key molecular features and structural motifs crucial for TRK inhibitor efficacy.
- To assess model generalization ability using different dataset partitioning strategies.
Main Methods:
- Utilized various algorithms to build descriptor-based and non-descriptor-based predictive models.
- Employed outer 10-fold cross-validation and random/cluster splitting for model evaluation.
- Combined deep neural network (DNN) with extended connectivity fingerprints 4 (ECFP4) descriptors.
- Performed structure-activity relationship (SAR) analysis using clustering and decision tree models.
Main Results:
- The DNN-ECFP4 model demonstrated excellent performance across random and cluster-split datasets.
- The DNN algorithm exhibited strong generalization prediction ability.
- Nitrogen-containing aromatic heterocyclic and benzo heterocyclic structures were identified as critical for TRK inhibitor activity.
Conclusions:
- The DNN-ECFP4 model is a powerful tool for predicting novel TRK inhibitors.
- Molecular feature richness is vital for predicting the activity of novel compounds.
- Specific heterocyclic structures are promising scaffolds for developing next-generation TRK inhibitors.
