Related Experiment Video
Updated: Oct 10, 2025
An Organotypic High Throughput System for Characterization of Drug Sensitivity of Primary Multiple Myeloma Cells
Published on: July 15, 2015
De novo Prediction of Cell-Drug Sensitivities Using Deep Learning-based Graph Regularized Matrix Factorization
Shuangxia Ren1, Yifeng Tao, Ke Yu
1Intelligent Systems Program, Pittsburgh, PA, USA.
Abstract:
Application of artificial intelligence (AI) in precision oncology typically involves predicting whether the cancer cells of a patient (previously unseen by AI models) will respond to any of a set of existing anticancer drugs, based on responses of previous training cell samples to those drugs. To expand the repertoire of anticancer drugs, AI has also been used to repurpose drugs that have not been tested in an anticancer setting, i.e., predicting the anticancer effects of a new drug on previously unseen cancer cells de novo. Here, we report a computational model that addresses both of the above tasks in a unified AI framework. Our model, referred to as deep learning-based graph regularized matrix factorization (DeepGRMF), integrates neural networks, graph models, and matrix-factorization techniques to utilize diverse information from drug chemical structures, their impact on cellular signaling systems, and cancer cell cellular states to predict cell response to drugs. DeepGRMF learns embeddings of drugs so that drugs sharing similar structures and mechanisms of action (MOAs) are closely related in the embedding space. Similarly, DeepGRMF also learns representation embeddings of cells such that cells sharing similar cellular states and drug responses are closely related. Evaluation of DeepGRMF and competing models on Genomics of Drug Sensitivity in Cancer (GDSC) and Cancer Cell Line Encyclopedia (CCLE) datasets show its superiority in prediction performance. Finally, we show that the model is capable of predicting effectiveness of a chemotherapy regimen on patient outcomes for the lung cancer patients in The Cancer Genome Atlas (TCGA) dataset*.
Insights
This study introduces DeepGRMF, an artificial intelligence (AI) model for precision oncology. DeepGRMF predicts patient cancer cell drug responses and identifies new anticancer drugs, improving treatment strategies.
Area of Science:
- Computational biology
- Artificial intelligence in oncology
- Drug discovery and repurposing
Background:
- Precision oncology leverages AI to predict patient cancer cell response to existing anticancer drugs.
- AI can also repurpose drugs for novel anticancer applications by predicting their efficacy on new cancer cell types.
- Existing methods often address these tasks separately, limiting comprehensive drug response prediction.
Purpose of the Study:
- To develop a unified artificial intelligence (AI) framework, DeepGRMF, for predicting anticancer drug efficacy.
- To integrate diverse data sources including drug chemical structures, cellular signaling pathways, and cancer cell states.
- To enhance drug response prediction for both known and novel anticancer agents in precision oncology.
Main Methods:
- Developed DeepGRMF, a computational model integrating neural networks, graph models, and matrix factorization.
- The model learns embeddings for drugs and cells, capturing similarities in chemical structure, mechanism of action, and cellular states.
- Utilized Genomics of Drug Sensitivity in Cancer (GDSC) and Cancer Cell Line Encyclopedia (CCLE) datasets for model training and evaluation.
Main Results:
- DeepGRMF demonstrated superior prediction performance compared to existing models on GDSC and CCLE datasets.
- The model successfully learned meaningful representations of drugs and cells in an embedding space.
- Validated the model's capability to predict chemotherapy regimen effectiveness on lung cancer patient outcomes using The Cancer Genome Atlas (TCGA) data.
Conclusions:
- DeepGRMF provides a unified AI framework for predicting drug response in precision oncology.
- The model's ability to integrate diverse biological and chemical data enhances prediction accuracy for known and novel drugs.
- This approach holds promise for accelerating drug discovery, repurposing, and personalizing cancer treatment strategies.
