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 lines 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. In this work, we propose a deep learning framework, Personalized DeepSynergy 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 presonses 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
Halil Ibrahim Kuru, A. Ercument Cicek, Oznur Tastan
Our paper is available at bioRxiv
- Python 3.7
- Numpy 1.18.1
- Scipy 1.4.1
- Pandas 1.0.1
- Tensorflow 2.1.0
- Tensorflow-gpu 2.1.0
- Scikit-Learn 0.22.1
- keras-metrics 1.1.0
- h5py 2.10.0
- cudnn 7.6.5 (for gpu support only)
Raw data of drug combinations are taken from DrugComb
Drug chemical features are calculated by Chemopy
RMA normalized E-MTAB-3610 untrated cell line gene expression data is downloaded from cancerrxgene
Patient synergy and sensitivity data is obtained from the Supplementary Files of He et al. He et al.
You can download preprocessed data from link, extract all files into input_data/
$ python main.py --saved-model-name pdsp.h5 --train-test-mode 1Download pretrained weights from link
$ python main.py --saved-model-name best_model.h5 --train-test-mode 0- Zagidullin, B., Aldahdooh, J., Zheng, S., Wang, W., Wang, Y., Saad, J., ... & Tang, J. (2019). DrugComb: an integrative cancer drug combination data portal. Nucleic acids research, 47(W1), W43-W51.
- CCao, D. S., Xu, Q. S., Hu, Q. N., & Liang, Y. Z. (2013). ChemoPy: freely available python package for computational biology and chemoinformatics. Bioinformatics, 29(8), 1092-1094.
- [dataset] Francesco Iorio (2015). Transcriptional Profiling of 1,000 human cancer cell lines, arrayexpress-repository, V1. https://www.ebi.ac.uk/arrayexpress/experiments/E-MTAB-3610.
- [patient dataset] He, L., Tang, J., Andersson, E.I., Timonen, S., Koschmieder, S., Wennerberg, K., Mustjoki, S., Aittokallio, T.: Patient-customized drug combination prediction and testing for t-cell prolymphocytic leukemia patients. Cancer research 78(9), 2407–2418 (2018)