From cell lines to cancer patients: personalized drug synergy prediction
Halil Ibrahim Kuru1, A Ercument Cicek1,2, Oznur Tastan3
1Department of Computer Engineering, Bilkent University, Ankara 06800, Turkey.
Motivation:
Combination drug therapies are effective treatments for cancer. However, the genetic heterogeneity of the patients and exponentially large space of drug pairings pose significant challenges for finding the right combination for a specific patient. Current in silico prediction methods can be instrumental in reducing the vast number of candidate drug combinations. However, existing powerful methods are trained with cancer cell line gene expression data, which limits their applicability in clinical settings. While synergy measurements on cell line models are available at large scale, patient-derived samples are too few to train a complex model. On the other hand, patient-specific single-drug response data are relatively more available.
Results:
In this work, we propose a deep learning framework, Personalized Deep Synergy Predictor (PDSP), that enables us to use the patient-specific single drug response data for customizing patient drug synergy predictions. PDSP is first trained to learn synergy scores of drug pairs and their single drug responses for a given cell line using drug structures and large scale cell line gene expression data. Then, the model is fine-tuned for patients with their patient gene expression data and associated single drug response measured on the patient ex vivo samples. In this study, we evaluate PDSP on data from three leukemia patients and observe that it improves the prediction accuracy by 27% compared to models trained on cancer cell line data.
Availability And Implementation:
PDSP is available at https://github.com/hikuru/PDSP.
Insights
This study introduces a deep learning framework for personalized cancer drug synergy prediction. The model leverages patient-specific data to improve combination therapy selection, enhancing prediction accuracy by 27%.
Area of Science:
- Computational biology
- Pharmacogenomics
- Machine learning in oncology
Background:
- Combination drug therapies are crucial for cancer treatment but face challenges due to patient genetic heterogeneity and the vast search space of drug pairings.
- Current in silico prediction methods often rely on cancer cell line data, limiting their clinical applicability.
- Patient-derived data for training complex models is scarce, though single-drug response data is more accessible.
Purpose of the Study:
- To develop a deep learning framework for personalized prediction of drug synergy in cancer patients.
- To enable the use of patient-specific single-drug response data for customized synergy predictions.
- To improve the accuracy of in silico drug combination predictions for clinical settings.
Main Methods:
- Proposed a deep learning framework, Personalized Deep Synergy Predictor (PDSP).
- Trained PDSP on drug structures and large-scale cell line gene expression data to learn synergy scores and single-drug responses.
- Fine-tuned the model using patient gene expression data and ex vivo single-drug response measurements for personalized predictions.
Main Results:
- Evaluated PDSP on data from three leukemia patients.
- Demonstrated a 27% improvement in prediction accuracy compared to models trained solely on cancer cell line data.
- Showcased the framework's ability to customize drug synergy predictions using patient-specific information.
Conclusions:
- The Personalized Deep Synergy Predictor (PDSP) effectively utilizes patient-specific data for improved drug synergy prediction.
- This approach addresses the limitations of traditional methods relying on cell line data, paving the way for more personalized cancer therapies.
- PDSP offers a promising computational tool for optimizing combination drug selection in clinical oncology.
Related Concept Videos
02:50Combination Therapies and Personalized Medicine
The combination of the drug acetazolamide and sulforaphane is a good example of combination therapy to treat cancer. The cells in the interior of a large tumor often die due to the hypoxic and...
02:57Targeted Cancer Therapies
There are several types of targeted therapies against...